Running a small project on Hetzner from Germany. Got the email this morning. Honestly, even after the increase their dedicated boxes are still absurdly cheap compared to what you'd pay at AWS or GCP for equivalent specs.
The real story here isn't Hetzner being greedy. It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax. I priced out a RAM upgrade for my home server last week. Same kit I bought 8 months ago for 90 EUR is now 400+. That's not normal market dynamics.
What worries me more is the second-order effects. Startups that would normally spin up cheap VPS instances to prototype and iterate now face meaningfully higher costs at the exact stage where every euro matters. The "just deploy it" culture that made European indie dev scene so productive was built on sub-10 EUR/month boxes. Those days might be over for a while.
> It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax.
DRAM is priced based on supply and demand, like every other market.
When demand goes up, the price goes up for everyone. It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
> That's not normal market dynamics.
This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
I do find it interesting that so many people think “market rate” means the opposite of what economics teaches, and that prices should stay stable and not change much when the economic conditions change.
I also find it interesting to read all of the “we shouldn’t let them…” takes in response to this situation. The DRAM market is international. Trying to restrict it in one country would just see the data centers get built in another country.
But... They're not wrong. That IS the market. Unrestricted, gloriously free market with its historically predictable outcomes - yay!
That's not where the interesting discussion is. The interesting discussion is with the notion that free unregulated markets are universally good and will naturally lead to positive outcomes because... I don't know, I'm personally not religious, but somebody here will help me :-).
Sorry, when you say "gloriously free market", do you mean whatever it takes EU, helicopter money (or, rewinding a decade, Greenspan put) US, or factory of the world China? :)
My point is that it's not a real market economy if the risk premium -- and in China's case, the exchange rate -- is rigged. And it has been, since the 90s.
EDIT: For clarity, I'm agreeing with you, since you were being facetious.
Commodities used to be proper free markets. Many suppliers and many buyers of a product that was the same regardless of the supplier.
This lead to low prices and/or differentiation with new products.
Most of these markets were too good, so in general we now have a few big companies buying up the lion share of the supply so they can set the price regardless. For example soy, just to name one
An example of unregulated market is where I come to your house and put a gun to your head and in exchange for not pulling the trigger you give me your various items of value.
While you are technically correct, you are neglecting that it would a be a bad idea, because in such a market I would likely answer the door with a shotgun or I would have an agreement with my other neighbor to shoot you if you come to my door brandishing.
This is actually also how global diplomacy works. Either have big guns or big friends.
I think you have gone in the end of the spectrum, in a sense that even a state law's are being broken, we are talking about rules in the market itself.
An unrelated market is an oxymoron. You could come to my house and put a gun to my head, but that's not a consensual trade. That's just thuggery; the point of a market is that both sides benefit from trade.
For markets to exist, property rights also need to be respected.
People love to say that but they own a very small percentage of housing in reality. What’s driving housing costs is also supply and demand. Especially supply, since we’re not allowed to build any houses in most places people want to live.
You’re still missing the key point: Hedge funds and REITs aren’t arbitrarily buying housing at any cost.
They are responding to the market. If they overbuy then they will lose money and have to sell at a loss, at which point you could snap up some good deals.
This is ridiculously oversimplified, because there is no real market in housing. It is illegal to build in all of the places people want to buy. The purchase of housing by hedge funds isn't a problem on its own, it's simply a symptom of the bigger problem of supply restrictions.
The funds themselves say in their financials that they view housing as profitable because of the various restrictions on supply in every desirable city. They explicitly say that if those restrictions were lifted they would not be able to make money in that business and they would exit.
Hedge funds don’t have as high of institutional ownership as you assume. It’s actually pretty small.
That said, nothing about the situation you described is at odds with “free market”. You’re describing the operation of a free market.
I think a lot of people want “free market” to mean the opposite: A highly restricted market where they are protected from any supply and demand inputs from anyone else. They just want cheap things and don’t want to compete with anyone.
There are two sides to a free market, though. In your example where a hedge fund comes in and buys your entire neighborhood, they would have to do so by outbidding everyone. This drives up the price. If it’s an economically irrational move you’d be smart to sell your house to them at an inflated rate, too! Then move back in when the prices crash down.
I should point out the relevance of my argument, is completely independent from the fact the reply to this questions of yours, is higher than zero.
So dont see this reply as a justification. Just as a note that you failed to do basic diligence on distortions that are well known. And as I said, that are not relevant to the analogy.
That article doesn't support your point. Only a small fraction of the homes in that area are actually owned by hedge funds. You should check the facts before commenting.
> It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
Curious on whether you will still hold your stance if OpenAI gets a taxpayer bailout. Even disregarding a bailout, they are already lobbying hard for tax credit expansion.
A government bailout of OpenAI would be a regressive redistribution of wealth, which is a horrendously poor use of government funds. But that has no bearing on the fact that calling high DRAM prices induced by high demand a “tax” stretches the meaning of the word beyond all recognition.
There are many horrible things in the world and we don’t need to label them all as a “tax.” Words have meanings and if we use them in an imprecise way, it obfuscates the truth.
> Trying to restrict it in one country would just see the data centers get built in another country.
I'm surprised this isn't already what's being done. Inference doesn't require super low latency with the client, and the population's support of AI (and especially data centers for it) is waning quickly. This feels like another ideal use case for outsourcing the stuff Americans don't want to see to somewhere that it'll be someone else's problem.
> This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
The problem is that demand is being propped up by speculative capital. The AI companies are a bubble that is suffocating productive parts of the market with the hording of capital which they're now using to also hoard hardware. All this without making money for data centres that aren't build yet, for a handwavy promise that an AGI will magically solve all the worlds problems.
This is not normal, and it is not good for the broader economy.
Yeah the dudes argument is bunk when we remember that openAI bought CAPACITY and not actual product. The market is also heavily manipulated by the big 3 players in the market.
OpenAI brazenly used their market position to create artificial scarcity. That's not normal market behavior. That's manipulation. And now we all suffer.
> When demand goes up, the price goes up for everyone. It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
> This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
You act like it's a competitive market. It's not the case.
It's an oligopoly with an extremely inelastic supply side.
The market is already completely broken and ineffective due to concentration and export controls. The actual response to a major demand shock should be investments to increase capacities but it's currently extremely limited because suppliers want to protect their margins and fear the market contracting again.
I think the usefulness of market dynamics is their ability to follow things like factory capacity, which are themselves hard to follow, not the other way around.
For things that aren't inherently limited in production
it is supposed to work both ways..
A key element is that China still acts as a block..
So Chinese firms have lost a big opportunity by
not making DDR4 yet aren't ready with DDR5. When
they are ready it will probably tank the market
which is less profitable than selling at high
prices with actual availability of something
the whole time.
> DRAM is priced based on supply and demand, like every other market.
Please don't explain it away like that - you are referring to the theoretical "ideal" market where a bunch of small companies compete with low margins to the benefit of the wider customer base. This is not what is happening. We have a couple of intrinsically worthless, LLM-whale companies, working literally to swallow and entshittify literally everything in their weird transhumanist/accelerationist/weirdo way. To add to the insult, the whole creation of artificial scarcity is almost a political construct, paid for with "monopoly-the-game-money" that these companies DO NOT EARN but instead BORROW based on vague and dishonest promises of achieving a "Country of PhDs in a datacenter"/"Pocket PhDs"/"AGI by 2025" (oops, now apparently by 2028 according to the OpenAI CEO). In their weird vision, as humans we should be merely cattle to be managed, not independent spirits with interest and aspirations. That ghoul Karpathy speaks about "ghost in the machine", overlooking the magnificence of the already existing "ghost in the machine" in the form of human beings. We should not have to swallow the increasingly crappier future these folks are insisting on pushing on all of us.
Can't agree more. We can also predict with some confidence that in a year or two, supply would have adjusted and ram will be cheaper in the long run. We benefit from the expanded demand even if the fact that it first lands as a shock is disruptive to prices.
What makes it manipulation? If 5 companies want to buy a quadrilion ram chips to build datacenters, why is this manipulation moreso than a million companies each wanting to buy 100 ram chips?
I think the problem is that both the buyers and producers are too large. Governments should not allow companies to become this big, because... <gestures broadly at everything>. If there were a thousand ram makers and a thousand datacenter builders, this particular problem would not exist.
But you can't just label any price evolution you dislike as "price manipulation".
It's certainly price manipulation, but not likely to be intended price manipulation. Your arguments are flawed but you have reached the right conclusion.
This is one of the many flaws of badly regulated markets.
(There are no free markets, and there is never perfect information, and people often behave remarkably irrationally for many reasons.)
>×If 5 companies want to buy a quadrilion ram chips to build datacenters, why is this manipulation moreso than a million companies each wanting to buy 100 ram chips?
Because they are 5 companies, especially when it can be shown they work in unison (formed a cartel)
Economic history is full of examples of demand shocks. This is not some unique situation that has never occurred before.
This is actually a clean commodity price spike because it’s specifically not for market manipulation or financial engineering. It’s because demand for this product really did explode overnight.
> This is actually a clean commodity price spike because it’s specifically not for market manipulation or financial engineering. It’s because demand for this product really did explode overnight.
Based on how the same 3 billion has been circiling between Anthropic, OpenAI, Nvidia, Google, Microsoft, Amazon, and a few other companies... I really doubt that this is the case, to be honest.
I think it's reasonable to distinguish which side drove this. RAM prices are going up but it's not engineered primarily by RAM manufacturers. They are naturally jumping on the bandwagon and responding, but they aren't the drivers. Of course, how they respond matters. They could make other choices. Over time we'll see how this goes because AI could cool and then RAM manufacturers end up in a spot where they choose to manipulate prices to keep them higher.
Tax is also an economic term, which is not what’s happening. Calling it a “tax on consumers” doesn’t make sense because any data centers buying RAM right now are also buying from the same global market.
If commenters just want to be outraged and throw words around then use whatever words you want, I suppose.
The problem is that OpenAI has cornered the market. Maybe they haven't crossed the legal line or more to the point no one in this corrupt and incompetent administration is going to prosecute them, but buying up 40% of a market which hasn't got any additional capacity is cornering by any measure.
So yes, this is not a normal market. Your claim of a functioning market is the same as saying my laptop, having lit on fire, is a functioning computer after having 10,000 volts applied across it.
"The real story here isn't Hetzner being greedy. It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax."
We might also have our aquifers depleted and our electricity prices skyrocket. But at least we see really great benefits, such as being able to script some side-project while unemployed due to AI.
Anyone who thinks that modern data centers don't evaporate their "recirculated FRESH water" straight into the ocean can safely have their opinions summarily discarded.
Data centers consume...a lot...of water by design, recirculated water, does not means no water consumption.
Water must be continuously added in evaporative cooling systems used by many data centers.
[1] - Cooling towers reject heat through evaporation, which uses water, not just recirculates it. Evaporated water is lost to the atmosphere and must be replaced with "make-up" water. As a result, recirculating cooling loops still require new water input to make up evaporation and blowdown losses.
But aren't those the same startups that think they need to run on AWS EKS instead of using a single cheap server? The cheapest used Hetzner server currently is €39.24 / month:
Similar to my favourite OVH servers, but I have unlimited traffic at 0.5Gb/s 64gb ram and dual mics. Similar price (with vat in Poland).
If you wanted to run same workloads on Aws it would cost you few hundred euro a month.
I see a silver lining to all this. At least maybe the silly "throw more horizontal scaling at it" will stop being a default response to all performance problems and people that are able to squeeze more processing out of the same hardware will be sought after again.
If your only need is a lot of bandwidth with very low server CPU use that’s fine.
That CPU is ancient, though. Over a decade old. That DRAM is 2-channel DDR3.
This could be a good deal for someone, but entrusting your startup’s operations to a 10 year old slow computer in Germany instead of using EKS would be an extremely short sighted move. A startup should be developing software and shipping it quickly to validate the market, not pinching pennies to save the equivalent of a couple hours of developer salary.
Right, for exactly that reason Hetzner offers brand new AX42 / EX63 servers with ECC memory and modern (Zen 4 / Arrow Lake) CPUs for just a little bit more.
I would guess that 99.9% of startups wouldn't notice the age of the CPU if they aren't in the business for CPU compute power.
Also, if you don't want to provision software systems, you probably shouldn't use Kubernetes at all. Both this and compute are niche businesses and neither would rent a budget server anyway.
No, that's actually a really good deal for dedicated hardware with those specs. For a project sized for hardware like that, the CPU is a lot less relevant than the RAM and storage and transfer.
Measuring CPUs by thread count and clock speed is not a good way to gauge performance. A current gen CPU would be several times faster than this old CPU.
Depending on workload, this old CPU might be as slow as a 2 thread or even 1 thread current gen server.
It does 8000 CPU marks with 4 cores. Sure Xeon 674X does 83641 with 28 cores. But show me where can you find it for less than 10 times the price? And with 320GB RAM, 10TB of NVMe SSD storage and 10 GBit/s of "unlimited" bandwidth
More than that, compare it to modern cloud CPUs. Epyc 9845 gets 153000 but that's with 160 cores / 320 threads. Per core it's under 1000 and 4 cores would be 3825 when the 11-year-old i7 is 8000.
Because those big systems are optimized for power efficiency. That Epyc is ~2.4W/core compared to ~16W/core for the old i7. It has a lower base clock and is Zen5c. If we cut the 8-core Ryzen 9850X3D's score in half, 4 Ryzen cores from the same generation but with a higher base clock and six times the L3 cache per core would be 20942. But it's also back up to 15W/core. The Epyc still has better performance per watt.
The newer cores are significantly more efficient. That doesn't mean they're unconditionally faster independent of all other variables.
> And with 320GB RAM, 10TB of NVMe SSD storage and 10 GBit/s of "unlimited" bandwidth
I think you’re talking about something else. The comment above was about a machine that didn’t have 10TB of storage, 320GB RAM, or unlimited bandwidth.
If you find 320GB of RAM and unlimited bandwidth for 40 Euro monthly then send it over!
> Except 40€ a month is extremely poor value for this CPU that's more than a decade old.
This is a rather baffling opinion to have. All cloud providers charge far more for a virtualized instance running on God knows what hardware. You are faced with a deal where you can run your software on bare metal, and you complain about... About what exactly?
Excuse me, but if the difference between 10 EUR per month and 14 eur per month is going to kill your startup, you probably shouldn't try to start it.
Might be time to think about using and creating less memory-hungry software.
Actually I disagree. I've killed projects because I've run out of time for them and didn't like them costing me £50 a month. If I'd been able to keep them going at £10 a month, I might have kept them going until I could get back to them. Sometimes startups fail just because the owners get distracted by life, and the project just needs more time.
2) why could they not just up the prices for new deployments, like they did with their dedicated servers? I think that would be fairer to existing customers
If you have a company, I can recommend leaseweb for cheap hosting. I host my personal stuff like my email and my ente.io instance there. They are cheaper than Hetzner (already before the new price increase) if you don't need managed k8s.
The first known example is the 6th century BC, where a greek philosopher cornered the market on olive oil presses, because he predicted a richt harvest via his knowledge of astronomy:
The post seems to indicate this is just for VPSs, which doesn't seem true, the email I just received from Hetzner mentions price increases for dedicated servers too.
The ones I'm affected by seemingly:
Product -> previous price -> New price as of 1 April 2026
EX42-NVMe (FSN1) -> € 49.65 -> € 51.13
AX41 (FSN1) -> € 49.73 -> € 51.22
AX41-NVMe (FSN1) -> € 49.73 -> € 51.18
Server Auction -> € 65.22 -> € 67.18
Still cheap compared to the performance + unmetered bandwidth, so I'm personally not super upset about it, my monthly bill in total goes up maybe 40-50 EUR in total, not that outrageous.
Seems it's because of increased cost of hardware, and they seemingly tried to avoid increasing the prices but they couldn't. From the email:
> The underlying causes of the increased costs are, among others, the exploding demand for AI-related computing power and for cloud services. In addition, raw material prices and production costs have also generally risen for manufacturers. The costs for RAM and SSDs especially have risen by a large amount. For example, the cost for DRAM memory has increased up to 500% since September 2025. And according to market researchers like TrendForce, this price trend will continue throughout the year.
> We have genuinely tried hard to optimize our costs and to prevent increasing our prices for as long as possible. But we can no longer compensate for the strain that it has placed on our operations. We want to continue to deliver quality products that meet both our standards and your expectations, so we must take this step.
I'm not particularly tied to Hetzner as an American using their Ashburn servers, so I figured maybe this price increase puts them a little closer to DigitalOcean's pricing. The pricing is still pretty heavily in Hetzner's favor though: the CCX23s that I use will be $39.99 USD after April 1, but the closest DO equivalent is $126 USD with a third of the disk space.
A significant part of this is probably just the hockey-stick growth in the price of memory we have seen in the past 6 months. Would be surprised if this wasn't impacting their bottom line for maintenance.
RAM increased the most, but also SSD and HDD prices increased significantly. And it seems there are also supply problems, so you can't even be sure if you get the components you want at higher prices.
There is another factor at play here: EU hosting providers that are not owned lock, stock & barrel are few and far between and Hetzner has a very nice sales representative in the White House.
Pretty sure they are implying that the actions of the current president/administration are causing people to re-evaluate US dependencies. I don't really understand the first half
That the USA is no longer seen as a stable partner for the long term and that Trump with his idiotic policies and tariffs is driving sales for the few EU hosting scale-ups that are not somehow owned by America.
This doesn't solve the issue that globalism caused. Europe doesn't make DRAM nor has the know-how to quickly bring factories online which usually take 10+ years.
We are tied to American economy and if AI companies start driving prices up not only DRAM but basically everything will become more expensive.
Building a factory is one thing, they can have 50 of them built, but that doesn't mean much if all 50 together amount to like 0.1% of company's output.
Once those factories scale up to 5%, then we can start considering that they've actually built a domestic supply, but that's a whole different goal than simply building the factories. Building factories is trivial. Making them output something is also "trivial". Scaling that up to a meaningful amount is a whole different, much harder goal to accomplish.
It looks like it's still a big difference between how the US and EU are responding to the chip supply wars. The US is actually building their own manufacturing capabilities domestically while the EU is apparently doing nothing, which is unfortunate.
Infineon is _opening_ its fab plant in Dresden this year which was supported by around 1bn euros from the EU equivalent of the CHIPS Act. They started building this fab in 2023, while TSMC, who started building its fab in the US right after covid just delayed the opening to 2027
The fab that Infineon is building is vastly smaller in scale, and their tech isn't really relevant to this discussion. For instance, it doesn't produce CPU/GPU microchips or DRAM. Also only 300mm wafer technology, which isn't competitive for anything except for some narrow industrial use-cases. Glad to see the EU is doing it, but it's a completely different thing.
Pretty much everyone is on 300 mm wafers for everything now, and has been for a while. Are you perhaps reading this as 300
nm process (which would usually be called 0.3 micron)?
Well first of all, the CHIPS Act was not "axed", it is federal law passed by an overwhelming bipartisan majority of the House and Senate. It would take a complete reversal of congress to repeal it and it's still very popular among both parties.
> Well first of all, the CHIPS Act was not "axed", it is federal law passed by an overwhelming bipartisan majority of the House and Senate. It would take a complete reversal of congress to repeal it and it's still very popular among both parties.
DOGE cut basically all staff from the CHIPS Program Office, congress passed the money but Trump is choosing to turn it into a slush-fund the admin spends on industrial policy (such as buying a stake in Intel). Wolfspeed went into bankruptcy in part because the admin delayed CHIPS funding agreed by the previous admin [1] (it's unclear whether they received the grant now that they have left it).
This is news to a lot of Americans! The 2022 CHIPS and Science Act is codified federal law. I think a lot of states (Arizona, Idaho, New York) would be very interested to learn that the funding for the infrastructure that they are already building has somehow gone poof.
> Currently, 100% of leading-edge DRAM production occurs overseas, primarily in East Asia.[0]
They make DRAM for cars, not computers, in the USA. They've promised they'll bring manufacturing onshore any time soon, which effectively means they'll wait until Trump forgets about it.
The newfound desire to move away from American cloud providers isn’t related to pricing, it’s about the perception of growing instability within the American government, the perception of deteriorating freedom of speech, and the perception of an increasingly non-neutral business environment.
E.g., if I’m running a business in the US and I don’t kiss Trump’s ring (and pay bribes), if he becomes dictator for life in 2028, all bets are off for my business.
Both the EU and USA import the majority of their computer equipment, and the USA is placing heavy and unpredictable tariffs on those goods. It’s hard to argue that a business should bet that data centers will be cheaper in the US than in the EU if Trump is the last democratically elected president.
The most stable places to do business in 2026 are probably the EU and China.
Yeah, also Hetzner is smart enough to realize that a lot of people are moving to them who are "Buy EU"-driven and are less price sensitive (certainly the most valuable ones are). Hopefully they can take these higher prices and further invest into the platform.
Here's to hoping the IOU purchase orders for RAM and SSDs get cancelled... Though I think folks are hedging that this will happen and limiting new suppliers.
What happens when an unstoppable force (building everything in Electron because hardware is cheap) meets an immovable object (oh no hardware is expensive now)?
Every RAM producer is stopping their consumer grade RAM production to provide ECC-RAM and VRAM now. Micron discontinued and closed down Crucial brand as a whole.
So, getting systems with higher RAM capacity is getting harder (from laptops to smartphones). So, for a couple of years, we need to stop using Electron so much and use what we have efficiently.
Data centers, esp. AI hyperscalers do not care about efficiency for now, because they can suffocate consumer-grade part of the hardware marketplace and get anything and everything they want. When their bubble pops, or the whole capacity ends, they need to learn to be efficient, too.
For reference, a well-optimized cluster runs at ~90% efficiency even though they have thousands of users. AI hyperscalers are not there. Maybe 60% efficient, at most. They waste a lot of resources to keep their momentum.
AMD's RYZEN already supports it. ASUStor's latest generation of NAS devices come with AMD x86_64 processors and ECC RAM as a standard, but ECC RAM in SODIMM format was not cheap, even when the RAM was cheap, either.
I know, but the density of the data is much less in human case.
IOW, humans still learn more effectively with less information, because there are innate mechanisms which process this data continuously and extract new meanings from the same data. This is part of both intelligence and consciousness.
Why is that so unbelievable? TypeScript isn't JavaScript, and while they have the same runtime, compiled TypeScript often don't look like how you'd solve the same problem in vanilla JS, where you'd leverage the dynamic typing rather than trying to work around it.
The TS code looks very different from the JS code (which obviously is the point), but given that, not hard to imagine they have different runtime characteristics, especially for people who don't understand the inside and outs of JavaScript itself, and have only learned TypeScript.
One thing to consider, is that with JavaScript you put it in a .js file, point a HTML page at it, and that's it.
TypeScript uses a ton more than that, which would impact the amount of energy usage too, not to mention everything running the package registries and what not. Not sure if this is why the difference is bigger, as I haven't read the paper myself :)
But if you do, please do share what you find out about their methodology.
This image comes from running the different versions of the benchmark games programs. Some of the difference between languages may actually be just algorithmic differences, and also those programs are in general not representative of most of the software that runs.
That, and also because rust compiler is a very good guardrail & feedback mechanism for AI. I made 3 little tools that I use for myself without knowing how to write a single rust line myself.
I can see that a reality but I am more comfortable using Golang as the language compared to rust given its fast compile times and I have found it to be much more easier to create no-dependices/fewer-dependencies project plus even though I wouldn't consider myself a master in golang, maybe mediocre, I feel much easier playing with golang than rust.
The resource consumption b/w rust and golang would be pretty minimal to figure out actually for most use cases imho.
Like the purchase price + increased cost? The thing is that these parties are sitting on billions and billions of investor money, they don't care that hardware is 400x as expensive. Which companies like nvidia have capitalized on a few years ago, they were already able to price their hardware at a 400% markup compared to pre-crypto times, and shift their focus from consumer graphics chips to datacenter compute chips, causing their revenue to go up 6x (if my interpretation of [0] is correct)
Markets only "figure things out" in a petri dish economy where:
1) There are no barriers to entry for competitors (e.g. protectionist tariffs, equal access to capital for everyone)
2) There are perfect substitutes available, so transitioning to a competitor is seamless and free (e.g. no requirement to store data in Country X, no vendor lock-in, no security compliance)
3) The industry is not a "natural monopoly" when only a handful of vendors can operate due to capital investment and national/global distribution required (see power utilities, telecoms, petrochemicals)
4) Profitability attracts competitors (won't happen because of #3), but heavy competition prevents abnormal profits from accumulating to a single player (happens because of #1, #2 and #3)
When markets don't figure things out, as is the case around the world, you get a tangled mess of market failures, government intervention and lobbying to neuter proposed interventions.
Markets are never perfect but over the course of history they are a pretty good mechanism to solve these type of problems. Not sure why we think taxing hyperscalers differently is the answer. Government usually does worse than the market when it comes to sorting it out.
My argument is not that market is perfect but that the alternatives are probably far worse, like a new tax on a specific group of companies.
We see that market is very irrational now and it can stay irrational for long enough to destroy everything we know in tech.
By the time market figures things out, you may no longer have services, and hardware that you use daily. When such amounts of stupid money are pumped into a single industry, even if all AI companies went out of business tomorrow, it's going to take years for things to go back to normal.
FWIW, I'm not advocating taxes, as I think that won't really do anything. I don't know what the solution is either.
> ...and it can stay irrational for long enough to destroy everything we know in tech.
Nah. For decades software engineers have been more expensive than the cost of buying the extra hardware needed for vastly inefficient software. There are orders of magnitude of inefficiency there. So there's a ton of slack in the world's software that can be taken up by software engineers while hardware is scarce, pushing back the date where there will really be a problem probably by decades more.
Of course software engineers will see a problem though, because they'll have to learn to to write efficient software again.
ie. "Great, but now make it work with less RAM" will be a thing again, instead of "It needs more RAM so order some as it's cheaper than your time to fix the code".
Sounds like hyperbole. Yes the world is connected yes we are seeing shortages, yes the market is imperfect and it lags but this is how things get fixed. Prices are sorted out, manufacturers make bets on long term capacity. Some will be losers, some will be winners.
My guess is that many of the current people saying "technology is over and no one will afford their own computer" might have been born after the previous shortages, so it's in reality their first shortage and they have no memories (nor interest reading about) the previous ones, that all eventually washed over, even if at those points there were also people claiming that "No one will have their own SSD in the future, because prices will always be super expensive for consumers from now on".
That's my hypothesis I spent a whole of 30 seconds thinking about anyways.
This is a different kind of shortage though. Previous ones were cyclical and caused by supply/demand mismatches or natural disasters. This one is structural. The manufacturers are actively choosing to prioritize AI because the margins are dramatically higher, and AI market has virtually unlimited money right now.
> eventually washed over
Eventually is doing a lot of heavy lifting here. Several years of constrained supply have real consequences for people and businesses. Hardware manufacturers are saying most of their capacity is already sold out to AI customers through 2026, and possibly even through 2027 and 2028, with the rest of the markets getting what's left over. This is a fundamentally different market dynamic.
How is that different from today? The scale might be different, but it's quite literally a "supply/demand mismatch" right now.
I don't think what we're seeing today can be described as "structural", at least because it's way too short to make such proclamations today, if it ossifies, then yeah maybe I'd agree with you, it's become structural.
> Several years of constrained supply have real consequences for people and businesses
Indeed, but lets see if it'll go as far as being "several years", the prices already stopped increasing, and supply still isn't planned to be expanded, if either of those changes you might have a point, but as of today it seems like an exaggeration.
Market is fixing it. Memory makers prioritized HBM and enterprise NAND, some, like Crucial, went out of consumer business entirely.
At the same time, the rational market is behaving rationally - they're not increasing production because they're fearing AI bubble could burst, leaving them with oversupply and expensive factories.
The market, apart from AI market, is behaving exactly as it's designed and as it should. But it doesn't mean outcome is good for everyone.
> it can stay irrational for long enough to destroy everything we know in tech
What does this even mean? I know people on the internet sometimes exaggerate, but I cannot even begin to find a more charitable meaning with this, what exactly will be "destroyed" in "tech" because of prices going up for a year or two?
Here's an easy experiment to conduct: look around the room at your home and count all the devices that have a CPU, RAM, SSD or HDD.
Then take a look at your bank statement to see what are the services you pay for monthly that also require the same hardware.
Now, imagine that these devices or services can no longer procure RAM, SSD or HDD. There's no more available supply for these components, because this is what's happening.
Would you still be able to have these devices if they all broke tomorrow? What about your hypothetical Backblaze subscription? Would you still be able to have an off-site backup?
> imagine that these devices or services can no longer procure RAM, SSD or HDD
Why would I imagine something so far out from what will realistically happen?
Again, a lot of doom and gloom over very unrealistic scenarios. Where are you even getting this from, YouTube channels?
Of course if there is no RAM or flash-storage at all available, eventually hardware will be unfeasible. But when we've experienced these sort of things before, it eventually restores to "normal" prices, and there absolutely nothing pointing to what we're experiencing now to get even worse, if anything it's already stabilized.
Yeah, which shows that Valve don't think "these devices or services can no longer procure RAM, SSD or HDD" is actually what'll happen in reality, because then they'd have to cancel the hardware fully. Instead, they're delaying it.
Yet here we are. Markets kept growing, some companies lost during the supply crunch but hopefully we came out generally stronger. So again the doom and gloom is hard to track. Maybe we hit a couple years of different supply crunches in tech, at some point if demand sustains companies will figure it out, optimize price, manufacturing lines etc.
What we're seeing is the natural conclusion of VC distortion in a market. There is so much money being pumped into AI speculatively now that it's hurting normal and sustainable businesses in other parts of the economy.
The solution might have to be mandatory rationing of some kind to avoid a situation where only a handful of AI giants are able to buy essential components. We can't just throw the rest of the economy under a bus to support the AI bubble for a few more months.
I'm working with a business right now that would like to buy some new servers for sensible, boring business reasons. It is having trouble because the prices from their normal suppliers are now extremely high - if the components are even available at all. This business has nothing to do with AI or Big Tech and yet it's at risk of being unable to continue normal operations in much the same way that a business would be affected if the phone networks were all switched off or the water supply to its office was cut. We regulate those industries because their continued reasonable operation is essential to make sure everyone else can continue to operate reasonably as well.
I'm seeing the same thing. I was consulting a group of people in my city that wanted to digitize massive load of old VHS tapes. No AI, no crazy tech, just standard, boring storage+network infrastructure.
I'm looking at the procurement sheet that I made for them a year ago. Half of the items are no longer available, while the other half became so expensive that we'd probably build 10 of such labs with these costs a year ago.
I'm also looking at my home NAS right now - I pray not even a plastic clip breaks inside, because I'd have to shut it down.
While these are still likely the first things that you'd think of being affected, I'm sure the effects are rippling through essentially every industry that utilizes these components in their supply chain. Which is probably - every industry nowadays?
I think that’s a massive stretch. What we are seeing is a new frontier in tech that nobody knows where it will land yet. Hyperscalers see a future where if they don’t build now that they might be left behind.
Absolutely VC money is flowing around but I think it’s unclear where the cards fall yet.
Not sure what you would regulate here. I hate the tripe that America and China are at war but I do think it’s not a great decision to stop the current work the west is doing as China is pushing full steam ahead.
I wish this comment can be on the absolute top of this page. This really is one of my frustrations with the AI bubble.
Fwiw, the days of creating an good ol' reliable hosting provider/Vps provider are over. I looked extensively into it one time out of curiosity but this would be one of the worst times in history to do that.
We would be sort of stuck with the options that we have right now and more and more shops in Lowend are even shutting down or raising prices with the sheer ram crisis and even HDD and storage crisis now.
A provider in LET had a post which said, "what should we providers do to deal with the ram shortage/ram prices"
These providers gave competition/had different unique features too to have chosen them but they were also incredibly price sensitive and the AI bubble blew the sensitivity by raising the prices almost 5 times or more. This would impact real businesses.
Thank you for creating this comment. I hope more people can read this. I genuinely just want this bubble to burst asap so that we can see a sense of rationality back within the market/the market functioning as expected without the immense irrationality/unpredictability of future.
another point is this, from my hosting provider idea, I shut it down. Why? because it literally makes 0 sense to start now, its postponed indefinitely untill the bubble bursts/ram prices are decreased.
How many other projects might be going through something similar. Gck1's comment next to mine also gives an example of a project whose value of cost increased 10 times.
How many of such projects would simply be unable to be built because of the ram inflation can't be underestimated imo.
and forget people who wish to game and many other things too. Basic comodities in the previous year or two feel like luxury now. All because of AI. It's insane.
Because this perfect version of capitalism you think exists, doesn't.
We live in a world with markets dominated by cartels of tech companies who don't play by the rules. Every other industry that impacts society in a negative way typically pays some sort of specialized tax to offset that, I don't know why these tech oligarchs shouldn't have too. It's wild how people just want to let them do whatever they want.
Everyone says we need to deregulate tech, and certain industries to get ahead of China.. Isn't it funny how their largely government controlled economy (to a degree) is annihilating the west on all fronts economically. We need far more regulation.
China will defeat the West solely because it regulates its billionaires, not the other way around like we have it in the West. And I hope so, the world is rooting for you China.
Way to put words in people mouths. Markets are imperfect but I do believe on average they are one of the better tools to solve supply and demand issues.
I don’t know who will come out winners but I do agree that China did well taking the playbook from Singapore and navigating their country through incredible amounts of growth. They are still facing depressing housing prices and deflation in other parts of the economy.
There are absolutely areas where markets breakdown, thinking problems where impacts are on longer horizons but for simple supply and demand like what we are seeing today, things will sort out in a couple years.
Reads a bit like the Paperclip Maximizer appearing way ahead of schedule? Implemented not as AI, but as emergent behavior in the ways of the financial class (that happens to be about AI, singularity and all that).
These changes are effective April 1st for existing and new customers. The price increase ratios are also different across product lines.
* Cloud (VMs): 38%
* Bare metal: 15%
* Memory add-on for bare metal: 575% (effective immediately)
It feels like memory add-on is intentionally set high to discourage customers from adding more memory.
AX102 (128 GB RAM) costs €124, AX162 (256 GB RAM) costs €244, but the 128 GB memory add-on alone costs €264. If we ignore the setup fee, it’s more cost-effective to provision additional servers instead of adding RAM to bare metal instances.
By the same time next year the prices likely gone down, although maybe not to the pre-increase, but surely much lower than currently. Putting it in my calendar to revisit this comment in a year :)
Well, if all the doomers and gloomers were correct that this is the end of hardware at home, we'd see the price continue to increase, and suppliers trying to ramp up production, even if it'd take long time.
The fact that it stabilized (at whatever price) and that suppliers aren't even thinking about ramping up production, should tell people that the doomers and gloomers were yet again over-reacting to things they don't fully understand themselves.
> Grocery prices have also stabilized but I’m still paying too much
I think that's a local problem, if you happen to live in a country that's trying to move over to isolationism rather than globalism as of late. In other modern countries the prices are also increasing, but at least following inflation somewhat so the increase doesn't seem as bad for us. Maybe at least yet? Who knows.
> Well, if all the doomers and gloomers were correct that this is the end of hardware at home, we'd see the price continue to increase, and suppliers trying to ramp up production, even if it'd take long time.
Ramping up production takes months and paying back the price to ramp up production takes years. Manufacturers have started investing in more production capacity but it'll take a while before supply can be sold off.
Based on interviews with industry professionals, I believe the forecast is that RAM prices will start going down again between August and the end of next year. Until then, prices will climb as stock depletes and RAM production is capped.
> Manufacturers have started investing in more production capacity
Where are you getting this from? Because that's not what I've seen, if anything the industry seems to lowering the production capacity, not increasing it.
And even if it takes years, if they thought it was a sustainable growth in demand, they'd at least be moving in that direction which again, doesn't seem to be happening.
> I believe the forecast is that RAM prices will start going down again between August and the end of next year. Until then, prices will climb as stock depletes and RAM production is capped.
Before today, we used to be able to order an AX162-R for €207 and add 128 GB of RAM for €46. Starting today, the same calculator provides €207 for an AX162-R (*) and €264 for the 128 GB RAM add-on. Sadly, HN doesn't let me upload screenshots.
(*) The price change for AX162-R machines is effective starting April 1st.
> These changes are effective April 1st for existing and new customers.
Checking today doesn't really indicate anything.
It's worth noting that the hardware price of RAM is up at least 550% yoy, so this was always going to happen as soon as their existing contracts had to be renewed
I really love that their notification email includes applicable price change for my specific servers.
The worst counter example of this was Mercedes sending me an email saying "the terms and conditions have been updated, please read them at this link". It linked to the 52 page document I was supposed to read through in its entirety and manually diff against previous! Good thing they started adding a change log in the emails after some customer push back.
My ISP sends me an SMS telling me there's work being done in my area. I have 3 different accounts with them in different citites. No idea which one they are talking about at any given time.
Hm the pricing increase stresses me out out less than the server shortages. My impulse reaction is to buy a few cheap cloud VPS instances even though I don't need them right now... Anyone have any wisdom to encourage/discourage this?
I just bought a Raspberry Pi 4 1 GB memory with aluminum case, aluminum NVME adapter, and a 64 GB SSD for about 80 euros. With microsd it’s even cheaper. 4 GB RAM would be about 120 euros.
The 1 GB RAM replaces one Forgejo runner that was in Hetzner. With €5 per month, I will earn this investment back in less than two years. After the price increase, this period will only shorten!
Doesn't seem to apply to older/deprecated gen instances. I've got a CX22 there for personal screw-around projects and it's the same £3.95/mo (pre-VAT) afaict. So maybe not much help to folks ordering new or running on the current gen as the older kit isn't something you can order now, but a small boon for us laggards.
I almost didn't see their email because it's sent by "notification" (notification@hetzner.com, no name set). Title is "Update on our pricing".
Anyway my increase is:
EX44 (HEL1) € 44.76 -> € 50.76
Not pleased especially as the reason for the increase for existing customers is a nebulous "The costs to operate our infrastructure has increased dramatically."
Western memory manufacturers decided to chase the AI bubble, abandoning the consumer and low-requirement markets entirely.
Chinese manufacturers are now capturing that entire segment with full vertical integration. When this bubble stabilizes, because it will (it's not going to grow to infinite), Western companies won't recapture those markets.
They've already ceded competitive advantage for the next decade. This is a structural shift, not a cyclical shortage.
It's another step in the transformation of Western industry that began in the '80s: the shift from real economy and human-centric production to financialized operations.
Your ISP will cut your account when you saturate the upstream pipe 24/7 for weeks on end... which will only happen if you host video.
And your home insurance will not know/care if you're operating a desktop-sized computer or even a single server (it is perfectly fine and expected a developer might bring an actual server home for troubleshooting). Home insurance only cares if you're running dozens of them.
This excuse "we need to raise prices because we have more demand" is BS. They should be truthful and say "we can increase prices and people will pay it because they want to be EU based"
To be honest for anything more serious than a personal Minecraft server hetzner has been beaten by ovh for ages (on bandwidth - you get all you can eat data limited by speed from ovh - for example 500mbit, instead of 20tb from hetzner).
For this reason hetzner is always a "backup DC" in my eyes and never the primary.
Also I heard they are extremely sensitive regarding abuse allegations so don't even think of hosting something someone may not like seeing...
They get a lot of hype, but there are many competitors worth looking at.
companies that haven't turned a profit are outbidding the rest of the economy for hardware. that's not a supply shortage, it's a subsidy funded by venture capital.
On one hand this is not good but predictable. I'm on longer-term commitments with OVH, so it will be interesting to see how they follow. I'm still keeping Hetzner on my shopping list, even with the increase the bare-metal offerings are within my budget, and now that prices have increases they should be stable for a while (also import for budget management).
I haven't received this email, and I have one x64 server that costs around 4 EUR/mo, and an ARM server that costs about 6 EUR/mo. I wonder if I'll still be affected by the price increase.
You might be right. Sorry. I should have said they haven't increased prices for existing dedicated servers since my direct experince is only there. Actually until about 3-4 years ago when the whole world went to shit, using a server for a year or two then upgrading to a better server for cheaper, was the norm. In that environment, you would naturally not have price increases.
I also don't think you're right that it never happened for the dedicated servers :) I'm only using Hetzner for dedicated servers, and found an email from 2022 where they mention price updates:
> Unfortunately, we are forced to increase the prices on these Server Auction models [...] old price 37.60 Euro -> 59.29 Euro, comes into effect 2022-03-03
Probably not for existing customers (their existing servers). I don't recall anything other than the IPv4 related increase in the past. I might be wrong of course as I've demonstrated already.
Yes, in 2022 I was an existing customers, and my server increased in price then, the server affected at that point went from 37.60 Euro to 59.29 Euro. Today that same server went from 65.22 to 67.18, so there is even more price increases seemingly between today and 2022 but I'm not finding exactly when that was.
My CCX13 (dedicated cores) went from 15€ to 20€ now. Looking at Netcup as alternative, more cores and more RAM for 12€ - anybody has experience with their root (kvm'ed) servers?
This is likely just the first wave. If this component hoarding by AI continues, and it likely will, at some point, it will be just OpenAI and Anthropic who can afford to have compute.
This has affected SSDs first, then RAM, then HDD and it doesn't look like even HDD manufacturers are going to increase production. So unless groups of people suddenly learn how to manufacture all of this hardware and open factories quickly, it's going to be a very fun next few years.
People have been predicting SaaS will die for all the wrong reasons. It's not that anyone can ship a SaaS clone by prompting an AI, it's that nobody is going to have access to the hardware required.
This mirrors the increased costs of people who already space + power in a DC, and want to buy new machines to fill their racks. Everybody is being hit.
Wow. That sucks. hcloud was great for ages and highly competitively priced.
Vultr may be a good alternative. If you want to search VPS prices across the 6 major clouds (gcloud, aws-cli, hcloud, az, doctl, and vultr-cli) I made a wrapper TUI that lets you search, sort, and rent VPS.
This is it. Hetzner has always been very price competitive in its existence. Given the private ownership, I din‘t expect this to be a sudden outburst of greed, but to actually reflect rising costs.
If a provider has higher margins, they may choose to eat some of the cost. But I would not expect that to be the case across the board
I feel like a huge selling point of Hetzner is that they're based in Europe, and they're themselves citing that as the reason for a huge uptick in sales and new users. In that context, I don't Vultr is a realistic alternative.
Obviously the US pushing absolutely everyone away and making EU and Europe the new enemy, so now we here want to reciprocate that and feel the need to move away from US infrastructure ASAP.
Personally I've been on a personal quest to minimize my usage of US-based services for many years already, but right now it's even part of the mainstream conversations, so seems to be ramping up, finally.
For clients, I just do what they wish to do, and a bunch of them want to move to European infrastructure because they've seen what can happen when you rely on US infrastructure today, and don't want it to happen to them. Only one so far cited regulatory reasons, and I think they were misinformed, but helped them out anyways with it.
Personally I do it because it's better aligned with what kind of future I want, and not wanting to support hyper-captalism environments anymore.
Cloud Act directly conflicts with GDPR. To really rub salt in the wound Trump overtly threatening to invade the EU (Greenland) basically turned the whole of Europe off seeing the US as a reliable ally. I don't think Americans have caught up to how much damage he has done to the image of the US amongst allies. They seem blissfully unaware of what's happening. Of course there are plenty of astute Americans who are aware but not the public at large.
This will be as a shockwave in web hosting industry, the same as it was with electricity price. There is nowhere to run. Everyone will increase their prices, unless hardware crysis ends up.
This is probably the best alternative provider for individuals that I could find, unless you're orchestrating a fleet of servers or something. Personally I'll wait out my next billing cycle with hetzner, as I expect other hosts to follow shortly.
They are solid and cheaper, but they don't offer the same level of control plane and API access as Hetzner that is really helpful when managing a larger number of servers.
It is, but more customers at a time of historically high component prices will do it. If you set your costs assuming every user's hardware is $1, and your customer base doubles when the hardware is $2, you're going to have to raise prices for everybody
The next set of hardware purchases will cost more than their last set of hardware purchases, and that's going to outweigh any labour economies of scale given just how many hardware components are in shortage this year.
If their growth had been in their projections in say 2024, they might have just been able to skip a round of hardware purchases, but the combination of growth meaning they must expand their hardware and hardware costs made this inevitable.
Can anybody predict this craze? The classical memory manufacturers are not yet adding additional manufacturing capacity. They learned this hard way in the past. That means, the demand is here to stay for years without typical bubble burst. Is this a point where Chinese companies will rise worldwide?
The massive DC overbuild matches demand, prices normalise somewhat in 3-5 years.
The massive DC overbuild does not match demand, prices tank in 3-5 years.
Third possibility: some approach like Taalas renders the current storyline meaningless. Would put 3 in 10 odds of this happening but I'd looove to see it.
Fourth: entire planet gets profoundly sick of emdashes, we all move back into caves and live in eternal gratitude of the moment humanity woke up to how little all of this really matters.
Hard to predict. If the bubble pops (NVIDIA and "circular economy", massive FAANG datacenter expansion plans, huge LLM training budgets) the markets will once again be flooded with components.
But, the shortages may very well continue into 2027, leading to some manufacturers going out of business and yet another massive redistribution of wealth.
BuyFromEU is the funniest subreddit there is right now. Unintentionally but still entertaining. EU has managed to paint itself into unenviable corner. I can't buy from EU even thought I want to because for physical goods - cross country shipping costs are prohibitive and for digital - they are either subpar, more expensive or both.
Try this as experiment - try to buy something like precision dowel pins from Poland or DOLD Mechatronik with shipping to Greece, Bulgaria or Romania vs the same thing from Aliexpress or Temu. Chinese costs are cheaper even if they have to fly here.
Not to mention that from July 1, 2026, the EU is abolishing the €150 duty-free threshold for non-EU shipments. This is specifically targeted at the flood of packages from marketplaces like Temu and Shein.
From July there will be a flat customs duty of €3 for small consignments. This fee applies per category. If your package contains items from different product groups (e.g., a shirt and a cable), you might pay the fee multiple times.
The Goal: To create fair competition for European retailers who can't compete with subsidized shipping and tax loopholes from massive non-EU sellers.
This will obviously have a knock-on effect for larger shipped items which are presumably subsidised at the bottom line by these parcels of fast-fashion and eWaste.
As someone that frequently buys low-cost second hand electronics from Japan, I am a little frustrated about the €3 per-category customs duty. That means a €80 package of various old game cartridges, retro handhelds, digital watches and collectables will now have another €12 to €24 on top of the 21% VAT and €6 handling fee. For an €80 package I am now looking at €15 for shipping and €34 to €46 in import cost. That kills a fun hobby.
The reverse is true too. The shipping costs from US -> EU are prohibitively high, oftentimes making ordering stuff from the US a no go. I think only Amazon (for certain products at least) charges somewhat reasonable shipping costs.
>Try this as experiment - try to buy something like precision dowel pins from Poland or DOLD Mechatronik with shipping to Greece, Bulgaria or Romania vs the same thing from Aliexpress or Temu. Chinese costs are cheaper even if they have to fly here.
This is an awful experiment. Only consumers care about delivery costs on deliveries like these, and what you're looking at are explicitly not goods aimed at consumers.
Okay. Then buy pizza oven from Italy and see how shipping costs are 60% of the price of the oven.
Anyway - you seems to misunderstand. If transporting something from Shenzhen to Franfurt is cheaper than transporting the same thing from Krakow to Thessaloniki - means that EU has fucked up royally in its main mission - to facilitate movement of goods. WE have ungodly patch of local carriers and courier companies and a lot of friction in every kind of intra eu goods movement.
> Then buy pizza oven from Italy and see how shipping costs are 60% of the price of the oven
Is it even physically possible to deliver at a significantly lower cost? Pizza ovens are both very large and very heavy, you can't fit many of them in a vehicle. They're also tricky to load and unload.
> If transporting something from Shenzhen to Franfurt is cheaper than transporting the same thing from Krakow to Thessaloniki - means that EU has fucked up
Ummmm. No.
It means the United Nations Universal Post Union international treaties which effectively provide China with subsidised postage TO THE WORLD (as China is a "developing country") needs urgently updating....... Some of the postage you pay to send parcels within the boarders of your own country is used to subsidise crap posted from China.
Running a small project on Hetzner from Germany. Got the email this morning. Honestly, even after the increase their dedicated boxes are still absurdly cheap compared to what you'd pay at AWS or GCP for equivalent specs.
The real story here isn't Hetzner being greedy. It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax. I priced out a RAM upgrade for my home server last week. Same kit I bought 8 months ago for 90 EUR is now 400+. That's not normal market dynamics.
What worries me more is the second-order effects. Startups that would normally spin up cheap VPS instances to prototype and iterate now face meaningfully higher costs at the exact stage where every euro matters. The "just deploy it" culture that made European indie dev scene so productive was built on sub-10 EUR/month boxes. Those days might be over for a while.
> It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax.
DRAM is priced based on supply and demand, like every other market.
When demand goes up, the price goes up for everyone. It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
> That's not normal market dynamics.
This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
I do find it interesting that so many people think “market rate” means the opposite of what economics teaches, and that prices should stay stable and not change much when the economic conditions change.
I also find it interesting to read all of the “we shouldn’t let them…” takes in response to this situation. The DRAM market is international. Trying to restrict it in one country would just see the data centers get built in another country.
Saying this is just the market...is like saying housing is a free market after hedge funds buy your entire neighborhood...
But... They're not wrong. That IS the market. Unrestricted, gloriously free market with its historically predictable outcomes - yay!
That's not where the interesting discussion is. The interesting discussion is with the notion that free unregulated markets are universally good and will naturally lead to positive outcomes because... I don't know, I'm personally not religious, but somebody here will help me :-).
Sorry, when you say "gloriously free market", do you mean whatever it takes EU, helicopter money (or, rewinding a decade, Greenspan put) US, or factory of the world China? :)
My point is that it's not a real market economy if the risk premium -- and in China's case, the exchange rate -- is rigged. And it has been, since the 90s.
EDIT: For clarity, I'm agreeing with you, since you were being facetious.
Commodities used to be proper free markets. Many suppliers and many buyers of a product that was the same regardless of the supplier.
This lead to low prices and/or differentiation with new products.
Most of these markets were too good, so in general we now have a few big companies buying up the lion share of the supply so they can set the price regardless. For example soy, just to name one
Where are these unregulated markets? Are you trading with your neighbor? If so, good, the more of that the better.
An example of unregulated market is where I come to your house and put a gun to your head and in exchange for not pulling the trigger you give me your various items of value.
That's one form of trading with your neighbor.
While you are technically correct, you are neglecting that it would a be a bad idea, because in such a market I would likely answer the door with a shotgun or I would have an agreement with my other neighbor to shoot you if you come to my door brandishing.
This is actually also how global diplomacy works. Either have big guns or big friends.
I think you have gone in the end of the spectrum, in a sense that even a state law's are being broken, we are talking about rules in the market itself.
An unrelated market is an oxymoron. You could come to my house and put a gun to my head, but that's not a consensual trade. That's just thuggery; the point of a market is that both sides benefit from trade.
For markets to exist, property rights also need to be respected.
But these markets are only "free" if you ignore the net $5.5 trillion the Fed has printed post GFC.
https://fred.stlouisfed.org/series/WALCL
People love to say that but they own a very small percentage of housing in reality. What’s driving housing costs is also supply and demand. Especially supply, since we’re not allowed to build any houses in most places people want to live.
Prices are decided at the margins. Having PE and REITs at every single table, even if their actual ownership is small as a %, makes huge differences.
You’re still missing the key point: Hedge funds and REITs aren’t arbitrarily buying housing at any cost.
They are responding to the market. If they overbuy then they will lose money and have to sell at a loss, at which point you could snap up some good deals.
This is ridiculously oversimplified, because there is no real market in housing. It is illegal to build in all of the places people want to buy. The purchase of housing by hedge funds isn't a problem on its own, it's simply a symptom of the bigger problem of supply restrictions.
The funds themselves say in their financials that they view housing as profitable because of the various restrictions on supply in every desirable city. They explicitly say that if those restrictions were lifted they would not be able to make money in that business and they would exit.
> not allowed to build any houses in most places people want to live
Or to convert them into apartment buildings
Read you response again. You are reinforcing my argument in case you did not notice...
Hedge funds don’t have as high of institutional ownership as you assume. It’s actually pretty small.
That said, nothing about the situation you described is at odds with “free market”. You’re describing the operation of a free market.
I think a lot of people want “free market” to mean the opposite: A highly restricted market where they are protected from any supply and demand inputs from anyone else. They just want cheap things and don’t want to compete with anyone.
There are two sides to a free market, though. In your example where a hedge fund comes in and buys your entire neighborhood, they would have to do so by outbidding everyone. This drives up the price. If it’s an economically irrational move you’d be smart to sell your house to them at an inflated rate, too! Then move back in when the prices crash down.
Are there regulatory barriers to entering the DRAM market like there are in housing?
Which neighborhoods are now entirely owned by hedge funds?
I should point out the relevance of my argument, is completely independent from the fact the reply to this questions of yours, is higher than zero.
So dont see this reply as a justification. Just as a note that you failed to do basic diligence on distortions that are well known. And as I said, that are not relevant to the analogy.
"When Wall Street Is Your Landlord" - https://www.theatlantic.com/technology/archive/2019/02/singl...
"In one Atlanta zip code, they bought almost 90 percent of the 7,500 homes sold between January 2011 and June 2012"
That article doesn't support your point. Only a small fraction of the homes in that area are actually owned by hedge funds. You should check the facts before commenting.
correct, both of those things are examples of free markets
Adam Smith had already clarified free market refers to a market free from all forms of economic privilege, monopolies and artificial scarcities.
You are confusing market outcome with market structure.
Good point, however we can't miss the history which shows that free market are not achievable, in some fields of economic is seen as "fairy tale".
But where is the monopoly here? Nobody has a monopoly on housing supply or DRAM production.
Nobody is making an artificial scarcity. They’re producing as much as they can.
You can buy the exact same DRAM that the data centers are bidding on.
There is absolutely artificial scarcity in the housing market, it is one of the biggest problems we face as a country
Incumbent owners preventing the construction of more supply to maintain their own property values is not "free market". This is very basic stuff
> It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
Curious on whether you will still hold your stance if OpenAI gets a taxpayer bailout. Even disregarding a bailout, they are already lobbying hard for tax credit expansion.
A government bailout of OpenAI would be a regressive redistribution of wealth, which is a horrendously poor use of government funds. But that has no bearing on the fact that calling high DRAM prices induced by high demand a “tax” stretches the meaning of the word beyond all recognition.
There are many horrible things in the world and we don’t need to label them all as a “tax.” Words have meanings and if we use them in an imprecise way, it obfuscates the truth.
> Trying to restrict it in one country would just see the data centers get built in another country.
I'm surprised this isn't already what's being done. Inference doesn't require super low latency with the client, and the population's support of AI (and especially data centers for it) is waning quickly. This feels like another ideal use case for outsourcing the stuff Americans don't want to see to somewhere that it'll be someone else's problem.
> Trying to restrict it in one country would just see the data centers get built in another country.
Sounds like not stressing the electricity infrastructure in Spain, to run inference for Facebook North American posts, should be seen as a positive...
> This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
The problem is that demand is being propped up by speculative capital. The AI companies are a bubble that is suffocating productive parts of the market with the hording of capital which they're now using to also hoard hardware. All this without making money for data centres that aren't build yet, for a handwavy promise that an AGI will magically solve all the worlds problems.
This is not normal, and it is not good for the broader economy.
Yeah the dudes argument is bunk when we remember that openAI bought CAPACITY and not actual product. The market is also heavily manipulated by the big 3 players in the market.
OpenAI brazenly used their market position to create artificial scarcity. That's not normal market behavior. That's manipulation. And now we all suffer.
> When demand goes up, the price goes up for everyone. It’s not a “tax” on the rest of us in any sense. There’s just a lot of demand everywhere.
> This is actually a textbook example of markets functioning in response to a demand shock where supply cannot be increased rapidly.
You act like it's a competitive market. It's not the case. It's an oligopoly with an extremely inelastic supply side.
The market is already completely broken and ineffective due to concentration and export controls. The actual response to a major demand shock should be investments to increase capacities but it's currently extremely limited because suppliers want to protect their margins and fear the market contracting again.
Factory capacity does not follow market dynamics easily
I think the usefulness of market dynamics is their ability to follow things like factory capacity, which are themselves hard to follow, not the other way around.
For things that aren't inherently limited in production it is supposed to work both ways..
A key element is that China still acts as a block.. So Chinese firms have lost a big opportunity by not making DDR4 yet aren't ready with DDR5. When they are ready it will probably tank the market which is less profitable than selling at high prices with actual availability of something the whole time.
> DRAM is priced based on supply and demand, like every other market.
Please don't explain it away like that - you are referring to the theoretical "ideal" market where a bunch of small companies compete with low margins to the benefit of the wider customer base. This is not what is happening. We have a couple of intrinsically worthless, LLM-whale companies, working literally to swallow and entshittify literally everything in their weird transhumanist/accelerationist/weirdo way. To add to the insult, the whole creation of artificial scarcity is almost a political construct, paid for with "monopoly-the-game-money" that these companies DO NOT EARN but instead BORROW based on vague and dishonest promises of achieving a "Country of PhDs in a datacenter"/"Pocket PhDs"/"AGI by 2025" (oops, now apparently by 2028 according to the OpenAI CEO). In their weird vision, as humans we should be merely cattle to be managed, not independent spirits with interest and aspirations. That ghoul Karpathy speaks about "ghost in the machine", overlooking the magnificence of the already existing "ghost in the machine" in the form of human beings. We should not have to swallow the increasingly crappier future these folks are insisting on pushing on all of us.
Can't agree more. We can also predict with some confidence that in a year or two, supply would have adjusted and ram will be cheaper in the long run. We benefit from the expanded demand even if the fact that it first lands as a shock is disruptive to prices.
If the last five or so years are to go by, we'll just have another pricing shock by then. So yay.
In many countries, its illegal to manipulate prices in bulk.
What makes it manipulation? If 5 companies want to buy a quadrilion ram chips to build datacenters, why is this manipulation moreso than a million companies each wanting to buy 100 ram chips?
I think the problem is that both the buyers and producers are too large. Governments should not allow companies to become this big, because... <gestures broadly at everything>. If there were a thousand ram makers and a thousand datacenter builders, this particular problem would not exist.
But you can't just label any price evolution you dislike as "price manipulation".
It's certainly price manipulation, but not likely to be intended price manipulation. Your arguments are flawed but you have reached the right conclusion.
This is one of the many flaws of badly regulated markets.
(There are no free markets, and there is never perfect information, and people often behave remarkably irrationally for many reasons.)
>What makes it manipulation?
Their size and the effect on the market they have
>×If 5 companies want to buy a quadrilion ram chips to build datacenters, why is this manipulation moreso than a million companies each wanting to buy 100 ram chips?
Because they are 5 companies, especially when it can be shown they work in unison (formed a cartel)
Most markets don't have three purchasers trying to corner the entire supply of one product.
I guess we’d call this oligopsony?
The monopsony (single buyer of a good) equivalent of an oligopoly?
Phew, it is a word, but not a highly studied one!
https://en.wikipedia.org/wiki/Oligopsony
Economic history is full of examples of demand shocks. This is not some unique situation that has never occurred before.
This is actually a clean commodity price spike because it’s specifically not for market manipulation or financial engineering. It’s because demand for this product really did explode overnight.
> This is actually a clean commodity price spike because it’s specifically not for market manipulation or financial engineering. It’s because demand for this product really did explode overnight.
Based on how the same 3 billion has been circiling between Anthropic, OpenAI, Nvidia, Google, Microsoft, Amazon, and a few other companies... I really doubt that this is the case, to be honest.
I think it's reasonable to distinguish which side drove this. RAM prices are going up but it's not engineered primarily by RAM manufacturers. They are naturally jumping on the bandwagon and responding, but they aren't the drivers. Of course, how they respond matters. They could make other choices. Over time we'll see how this goes because AI could cool and then RAM manufacturers end up in a spot where they choose to manipulate prices to keep them higher.
It is manipulation when wafers are purchased in order to not process further a la OpenAI.
You're being exceedingly pedantic over the use of the colloquialism "DRAM tax" but then you allow "demand shocks". So yeah, everyone's shocked.
Weird hill..
Demand shock is not a colloquialism. It’s an real economic term that describes the situation.
https://en.wikipedia.org/wiki/Demand_shock
Tax is also an economic term, which is not what’s happening. Calling it a “tax on consumers” doesn’t make sense because any data centers buying RAM right now are also buying from the same global market.
If commenters just want to be outraged and throw words around then use whatever words you want, I suppose.
By 3 buyers who have no known plan to finance the purchase orders they have made.
Economic history is also full of examples of bubble bursts.
The problem is that OpenAI has cornered the market. Maybe they haven't crossed the legal line or more to the point no one in this corrupt and incompetent administration is going to prosecute them, but buying up 40% of a market which hasn't got any additional capacity is cornering by any measure.
So yes, this is not a normal market. Your claim of a functioning market is the same as saying my laptop, having lit on fire, is a functioning computer after having 10,000 volts applied across it.
AI demand is subsidized by the bubble. Those operators buying the RAM are not paying using money that exists. Market economics are not working here.
"it happened therefore it's normal"
"The real story here isn't Hetzner being greedy. It's that AI companies are vacuuming up every DRAM chip on the planet and the rest of us get to pay the tax."
We might also have our aquifers depleted and our electricity prices skyrocket. But at least we see really great benefits, such as being able to script some side-project while unemployed due to AI.
>aquifers depleted
Oh it's this thoroughly debunked talking point again.
https://andymasley.substack.com/p/the-ai-water-issue-is-fake
Anyone who thinks modern data centers don’t use recirculated water can safely have their opinions summarily discarded.
Please google "datacenter evaporative cooling" and then re-evaluate
Anyone who thinks that modern data centers don't evaporate their "recirculated FRESH water" straight into the ocean can safely have their opinions summarily discarded.
Data centers consume...a lot...of water by design, recirculated water, does not means no water consumption. Water must be continuously added in evaporative cooling systems used by many data centers.
[1] - Cooling towers reject heat through evaporation, which uses water, not just recirculates it. Evaporated water is lost to the atmosphere and must be replaced with "make-up" water. As a result, recirculating cooling loops still require new water input to make up evaporation and blowdown losses.
[1] - https://en.wikipedia.org/wiki/Cooling_tower
But aren't those the same startups that think they need to run on AWS EKS instead of using a single cheap server? The cheapest used Hetzner server currently is €39.24 / month:
- Intel Core i7-6700 - 32 GB - 2 x 480 GB Datacenter SSD - 1 GB/s - 20 TB traffic
Their VPS are even cheaper. And you can run a lot on this.
Similar to my favourite OVH servers, but I have unlimited traffic at 0.5Gb/s 64gb ram and dual mics. Similar price (with vat in Poland).
If you wanted to run same workloads on Aws it would cost you few hundred euro a month.
I see a silver lining to all this. At least maybe the silly "throw more horizontal scaling at it" will stop being a default response to all performance problems and people that are able to squeeze more processing out of the same hardware will be sought after again.
If your only need is a lot of bandwidth with very low server CPU use that’s fine.
That CPU is ancient, though. Over a decade old. That DRAM is 2-channel DDR3.
This could be a good deal for someone, but entrusting your startup’s operations to a 10 year old slow computer in Germany instead of using EKS would be an extremely short sighted move. A startup should be developing software and shipping it quickly to validate the market, not pinching pennies to save the equivalent of a couple hours of developer salary.
>That CPU is ancient, though. Over a decade old.
Coincidentally so are the t3 / t3a instances on AWS that everyone loves to use especially for dev/staging environments
10 years is where hardware failure rates start ramping quickly, in my experience.
Not necessarily obvious failures, but subtle errors, memory problems (like this case without an ECC capable CPU) and little instabilities.
With cloud instances I can migrate to a new instance with a couple clicks if I want.
Trying to save a couple hundred euros per month on hosting costs needs to be balanced against the risks and extra developer time.
For personal projects these old instances can be an excellent deal though.
Right, for exactly that reason Hetzner offers brand new AX42 / EX63 servers with ECC memory and modern (Zen 4 / Arrow Lake) CPUs for just a little bit more.
I said the same thing to myself.
But then I remembered that what AWS gives you is the same generation of CPU, just obfuscated.
GCP Also obfuscates it, but not as much: https://docs.cloud.google.com/compute/docs/general-purpose-m...
(note: skylake is 10 years old)
> That CPU is ancient, though. Over a decade old. That DRAM is 2-channel DDR3.
6700 should be DDR4 unless they're using some weird-ass setup.
I would guess that 99.9% of startups wouldn't notice the age of the CPU if they aren't in the business for CPU compute power.
Also, if you don't want to provision software systems, you probably shouldn't use Kubernetes at all. Both this and compute are niche businesses and neither would rent a budget server anyway.
While I agree with the last sentence, I would suggest you buy what is needed not what is latest.
Except 40€ a month is extremely poor value for this CPU that's more than a decade old.
Is it? It's 10x the price on GCP:
https://cloud.google.com/products/calculator?dl=CjhDaVEyWWpJ...
The CPU is the same generation; https://docs.cloud.google.com/compute/docs/general-purpose-m...
No, that's actually a really good deal for dedicated hardware with those specs. For a project sized for hardware like that, the CPU is a lot less relevant than the RAM and storage and transfer.
For the 5 api requests a second most projects will get, it'll probably do.
8 threads at 3.4 GHz, 8MB cache. Seems fine, depending on your use case.
Measuring CPUs by thread count and clock speed is not a good way to gauge performance. A current gen CPU would be several times faster than this old CPU.
Depending on workload, this old CPU might be as slow as a 2 thread or even 1 thread current gen server.
It does 8000 CPU marks with 4 cores. Sure Xeon 674X does 83641 with 28 cores. But show me where can you find it for less than 10 times the price? And with 320GB RAM, 10TB of NVMe SSD storage and 10 GBit/s of "unlimited" bandwidth
More than that, compare it to modern cloud CPUs. Epyc 9845 gets 153000 but that's with 160 cores / 320 threads. Per core it's under 1000 and 4 cores would be 3825 when the 11-year-old i7 is 8000.
Because those big systems are optimized for power efficiency. That Epyc is ~2.4W/core compared to ~16W/core for the old i7. It has a lower base clock and is Zen5c. If we cut the 8-core Ryzen 9850X3D's score in half, 4 Ryzen cores from the same generation but with a higher base clock and six times the L3 cache per core would be 20942. But it's also back up to 15W/core. The Epyc still has better performance per watt.
The newer cores are significantly more efficient. That doesn't mean they're unconditionally faster independent of all other variables.
> And with 320GB RAM, 10TB of NVMe SSD storage and 10 GBit/s of "unlimited" bandwidth
I think you’re talking about something else. The comment above was about a machine that didn’t have 10TB of storage, 320GB RAM, or unlimited bandwidth.
If you find 320GB of RAM and unlimited bandwidth for 40 Euro monthly then send it over!
Yes, e.g. for AWS it pays off to have a look at the 'CoreMark Score' column at https://instances.vantage.sh/
If you need more power check out the AX line of dedicated servers: https://www.hetzner.com/dedicated-rootserver/matrix-ax/
Except you're getting a couple of disks, many GB of RAM, and some on-site 24/7 support, limitless network traffic, and your electricity bill.
Not too bad considering.
> Except 40€ a month is extremely poor value for this CPU that's more than a decade old.
This is a rather baffling opinion to have. All cloud providers charge far more for a virtualized instance running on God knows what hardware. You are faced with a deal where you can run your software on bare metal, and you complain about... About what exactly?
Excuse me, but if the difference between 10 EUR per month and 14 eur per month is going to kill your startup, you probably shouldn't try to start it. Might be time to think about using and creating less memory-hungry software.
Actually I disagree. I've killed projects because I've run out of time for them and didn't like them costing me £50 a month. If I'd been able to keep them going at £10 a month, I might have kept them going until I could get back to them. Sometimes startups fail just because the owners get distracted by life, and the project just needs more time.
1) this reads like it's posted by an LLM
2) why could they not just up the prices for new deployments, like they did with their dedicated servers? I think that would be fairer to existing customers
If you have a company, I can recommend leaseweb for cheap hosting. I host my personal stuff like my email and my ente.io instance there. They are cheaper than Hetzner (already before the new price increase) if you don't need managed k8s.
Hetzner should not be compared to AWS or GCP for pricing. It should be compared to Vultr, Linode or DigitalOcean.
Of which DigitalOcean is running very outdated prices at 4USD for 512mb ram.
The cheapest Kimsufi dedicated server with 32GB ram is $11.10/mo.
On the plus siide, we all get to learn how few new computers are needed rather than chase number goes up.
> Same kit I bought 8 months ago for 90 EUR is now 400+. That's not normal market dynamics.
That's exactly normal market dynamics during acute shortage. Remember 2020 when filtering face masks went up in price 10-100x?
Warning: another bot - check their history.
The reason Hetzner was cheap was bad latency and Arbor.
> That's not normal market dynamics.
It is, in fact, normal market dynamics.
It's normal market dynamics for markets dominated by quasi-monopolies, which is why regulation should have prevented the existence of such markets.
I don't buy that it's because of the monopoly. TSMC has been starting a fab for close to 5 years.
It doesn't matter how many companies are in this market, it takes a real amount of time to add capacity.
A single company that never made a profit outbiding the entire world is normal?
Yes. In fact that's not just normal, that's very frequent in market.
Please provide historical examples.
The first known example is the 6th century BC, where a greek philosopher cornered the market on olive oil presses, because he predicted a richt harvest via his knowledge of astronomy:
https://en.wikipedia.org/wiki/Cornering_the_market#Thales_of...
Maybe not normal market dynamics, but typical human behavior: https://en.wikipedia.org/wiki/Vincent_Kosuga#Cornering_the_o...
The post seems to indicate this is just for VPSs, which doesn't seem true, the email I just received from Hetzner mentions price increases for dedicated servers too.
The ones I'm affected by seemingly:
Still cheap compared to the performance + unmetered bandwidth, so I'm personally not super upset about it, my monthly bill in total goes up maybe 40-50 EUR in total, not that outrageous.Here is the full list of the updated prices: https://docs.hetzner.com/general/infrastructure-and-availabi...
Seems it's because of increased cost of hardware, and they seemingly tried to avoid increasing the prices but they couldn't. From the email:
> The underlying causes of the increased costs are, among others, the exploding demand for AI-related computing power and for cloud services. In addition, raw material prices and production costs have also generally risen for manufacturers. The costs for RAM and SSDs especially have risen by a large amount. For example, the cost for DRAM memory has increased up to 500% since September 2025. And according to market researchers like TrendForce, this price trend will continue throughout the year.
> We have genuinely tried hard to optimize our costs and to prevent increasing our prices for as long as possible. But we can no longer compensate for the strain that it has placed on our operations. We want to continue to deliver quality products that meet both our standards and your expectations, so we must take this step.
I'm not particularly tied to Hetzner as an American using their Ashburn servers, so I figured maybe this price increase puts them a little closer to DigitalOcean's pricing. The pricing is still pretty heavily in Hetzner's favor though: the CCX23s that I use will be $39.99 USD after April 1, but the closest DO equivalent is $126 USD with a third of the disk space.
A significant part of this is probably just the hockey-stick growth in the price of memory we have seen in the past 6 months. Would be surprised if this wasn't impacting their bottom line for maintenance.
RAM increased the most, but also SSD and HDD prices increased significantly. And it seems there are also supply problems, so you can't even be sure if you get the components you want at higher prices.
There is another factor at play here: EU hosting providers that are not owned lock, stock & barrel are few and far between and Hetzner has a very nice sales representative in the White House.
Can you expound on that? I'm not sure I get what you're implying.
Pretty sure they are implying that the actions of the current president/administration are causing people to re-evaluate US dependencies. I don't really understand the first half
I think in the first part they are implying that there are very few independent companies to turn to.
(I also prefer comments that are clear without insinuations).
Precisely like code
What about all of the long-tail providers that are often listed on lowendbox.com and similar sites?
Ahh, the sales rep is Trump, that makes sense, thank you. I thought Jacques meant they had lobbyists somehow.
1. There's no meaningful European competition.
2. Trump is making everyone scared to use US hosting.
So they're leveraging for extra profits.
That the USA is no longer seen as a stable partner for the long term and that Trump with his idiotic policies and tariffs is driving sales for the few EU hosting scale-ups that are not somehow owned by America.
That Trump makes us very motivated to stop relying on American tech.
This doesn't solve the issue that globalism caused. Europe doesn't make DRAM nor has the know-how to quickly bring factories online which usually take 10+ years.
We are tied to American economy and if AI companies start driving prices up not only DRAM but basically everything will become more expensive.
America doesn't manufacture DRAM either, this is all South Korea and Taiwan.
??? Micron has DRAM megafabs in both Idaho and New York state.
https://en.wikipedia.org/wiki/Micron_Technology
They don't "have them", they're building them.
https://www.micron.com/us-expansion/id
> Micron has already achieved key construction milestones on its first Idaho fab with DRAM output scheduled to begin in 2027.
https://investors.micron.com/news-releases/news-release-deta...
> Production is expected to start in 2030 with the fabs ramping throughout the decade.
Until they start outputting DRAM in any meaningful quantity, they're not relevant.
> They don't "have them", they're building them.
According to wikipedia Micron Fab 6 in Virginia started production in 1997 and is still operating
> "in any meaningful capacity"
Building a factory is one thing, they can have 50 of them built, but that doesn't mean much if all 50 together amount to like 0.1% of company's output.
Once those factories scale up to 5%, then we can start considering that they've actually built a domestic supply, but that's a whole different goal than simply building the factories. Building factories is trivial. Making them output something is also "trivial". Scaling that up to a meaningful amount is a whole different, much harder goal to accomplish.
It looks like it's still a big difference between how the US and EU are responding to the chip supply wars. The US is actually building their own manufacturing capabilities domestically while the EU is apparently doing nothing, which is unfortunate.
Infineon is _opening_ its fab plant in Dresden this year which was supported by around 1bn euros from the EU equivalent of the CHIPS Act. They started building this fab in 2023, while TSMC, who started building its fab in the US right after covid just delayed the opening to 2027
The fab that Infineon is building is vastly smaller in scale, and their tech isn't really relevant to this discussion. For instance, it doesn't produce CPU/GPU microchips or DRAM. Also only 300mm wafer technology, which isn't competitive for anything except for some narrow industrial use-cases. Glad to see the EU is doing it, but it's a completely different thing.
Pretty much everyone is on 300 mm wafers for everything now, and has been for a while. Are you perhaps reading this as 300 nm process (which would usually be called 0.3 micron)?
Are those plants still functional after CHIPS act was axed? I thought they mainly produce in Asia now.
Well first of all, the CHIPS Act was not "axed", it is federal law passed by an overwhelming bipartisan majority of the House and Senate. It would take a complete reversal of congress to repeal it and it's still very popular among both parties.
Where do you get your information from?
> Well first of all, the CHIPS Act was not "axed", it is federal law passed by an overwhelming bipartisan majority of the House and Senate. It would take a complete reversal of congress to repeal it and it's still very popular among both parties.
DOGE cut basically all staff from the CHIPS Program Office, congress passed the money but Trump is choosing to turn it into a slush-fund the admin spends on industrial policy (such as buying a stake in Intel). Wolfspeed went into bankruptcy in part because the admin delayed CHIPS funding agreed by the previous admin [1] (it's unclear whether they received the grant now that they have left it).
[1] https://www.ft.com/content/4aac09f9-19df-401a-9ab3-ef14a47bb...
Once again, where do you get your information from? The only thing that doesn't exist anymore is DOGE itself.
https://wtop.com/government/2025/11/doge-quietly-disbands-8-...
> after CHIPS act was axed
This is news to a lot of Americans! The 2022 CHIPS and Science Act is codified federal law. I think a lot of states (Arizona, Idaho, New York) would be very interested to learn that the funding for the infrastructure that they are already building has somehow gone poof.
> Currently, 100% of leading-edge DRAM production occurs overseas, primarily in East Asia.[0]
They make DRAM for cars, not computers, in the USA. They've promised they'll bring manufacturing onshore any time soon, which effectively means they'll wait until Trump forgets about it.
0: https://www.nist.gov/news-events/news/2025/06/president-trum...
That's not how it works, DRAM substrate (the actual chip that contains memory cells) is shared. It's only the packaging that differs.
And China with IXMT.
Europe has stopped making DRAM relatively recently (Qimonda).
This should have not been allowed to happen.
Isn’t there also basically 0 American DRAM?
Micron Technology, Inc. is an American semiconductor company that manufactures computer memory
They don't produce them within the US. They're building some factories to do so in the future, but as of now their output is 0.
The newfound desire to move away from American cloud providers isn’t related to pricing, it’s about the perception of growing instability within the American government, the perception of deteriorating freedom of speech, and the perception of an increasingly non-neutral business environment.
E.g., if I’m running a business in the US and I don’t kiss Trump’s ring (and pay bribes), if he becomes dictator for life in 2028, all bets are off for my business.
Both the EU and USA import the majority of their computer equipment, and the USA is placing heavy and unpredictable tariffs on those goods. It’s hard to argue that a business should bet that data centers will be cheaper in the US than in the EU if Trump is the last democratically elected president.
The most stable places to do business in 2026 are probably the EU and China.
You waited far too long.
Yeah, also Hetzner is smart enough to realize that a lot of people are moving to them who are "Buy EU"-driven and are less price sensitive (certainly the most valuable ones are). Hopefully they can take these higher prices and further invest into the platform.
The strongest reaction of EU would be to subsidize RIPE small LIR fees to 0€ and embrace decentralization.
What would that achieve? Here, have 1.5% discount on your subnet purchase
Here's to hoping the IOU purchase orders for RAM and SSDs get cancelled... Though I think folks are hedging that this will happen and limiting new suppliers.
It seems we will run out of hardware by March?
"Hard drives already sold out for this year" - https://www.theregister.com/2026/02/20/ai_blamed_again_as_ha...
Time for an AI tax on the hyperscalers.
> It seems we will run out of hardware by March?
What happens when an unstoppable force (building everything in Electron because hardware is cheap) meets an immovable object (oh no hardware is expensive now)?
We go back to the demoscene days, being creative with what we have instead of shipping Electron junk.
Inshallah
If the FSM wills
Maybe we need to let go of our auto-scaled 100 pod service mesh for a todo list app, and just deploy it bare metal on 2 servers.
consumer RAM is not what's creating shortage. Data centers doesn't run electron to train the model or for inference
Sure, consumer ram isn't causing a shortage, but it's affected by the shortage.
Every RAM producer is stopping their consumer grade RAM production to provide ECC-RAM and VRAM now. Micron discontinued and closed down Crucial brand as a whole.
So, getting systems with higher RAM capacity is getting harder (from laptops to smartphones). So, for a couple of years, we need to stop using Electron so much and use what we have efficiently.
Data centers, esp. AI hyperscalers do not care about efficiency for now, because they can suffocate consumer-grade part of the hardware marketplace and get anything and everything they want. When their bubble pops, or the whole capacity ends, they need to learn to be efficient, too.
For reference, a well-optimized cluster runs at ~90% efficiency even though they have thousands of users. AI hyperscalers are not there. Maybe 60% efficient, at most. They waste a lot of resources to keep their momentum.
I have a silent hope that because of this change we all will get ECC ram and that consumer CPUs will get proper support for them.
AMD's RYZEN already supports it. ASUStor's latest generation of NAS devices come with AMD x86_64 processors and ECC RAM as a standard, but ECC RAM in SODIMM format was not cheap, even when the RAM was cheap, either.
They effectively do. They’re trained by brute forcing 100TB of training data through them, rather than any logical learning technique.
A human doesn’t need 100TB of books to learn the alphabet.
> A human doesn’t need 100TB of books to learn the alphabet.
A human does need 16ish hours per day of audio/video content for several years to learn the alphabet.
I used a single letter stencil to learn the alphabet, actually, and nobody strapped me to a chair to watch or listen something 16h a day.
Living inside a normal home with my parents was enough for the audio part.
The 16 hours of audio/video per day was a reference to being alive and hearing/seeing things for years before you actually could learn the alphabet.
It was not meant as literally sitting at a screen with audio/video for 16 hours a day.
I know, but the density of the data is much less in human case.
IOW, humans still learn more effectively with less information, because there are innate mechanisms which process this data continuously and extract new meanings from the same data. This is part of both intelligence and consciousness.
LLMs lack both.
> humans still learn more effectively with less information
> because there are innate mechanisms which process this data continuously and extract new meanings from the same data
To me, these statements strongly contradict each other, but I also really do not care enough to debate it.
I respect your disagreement and desire to leave the debate here. So we can agree to disagree.
Have a nice day.
I guess we have to get creative again.
I actually think you're right here.
Resource constraints have often helped me come up with stuff that I'm actually proud of.
2026 will be the year of Rust...
Due to lack of memory leaks which will stop increasing RAM prices?
Because it's more memory efficient than most other languages. So you can achieve the same result with lower RAM requirements.
The efficiency...
https://users.rust-lang.org/t/energy-consumption-in-programm...
I see that's from almost 10 years ago, it would be interesting to see how that's changed with improvements to V8, python and C# since.
Also, Typescript 5 times worse than Javascript? That doesn't really make sense, since they share the same runtime.
Why is that so unbelievable? TypeScript isn't JavaScript, and while they have the same runtime, compiled TypeScript often don't look like how you'd solve the same problem in vanilla JS, where you'd leverage the dynamic typing rather than trying to work around it.
See this example as one demonstration: https://www.typescriptlang.org/play/?q=8#example/enums
The TS code looks very different from the JS code (which obviously is the point), but given that, not hard to imagine they have different runtime characteristics, especially for people who don't understand the inside and outs of JavaScript itself, and have only learned TypeScript.
Enums are one of only a few places where there is significant deviation, I don't believe that makes it 400% less efficient.
Maybe read the paper and see if you can figure out their reasoning/motivation :) https://dl.acm.org/doi/10.1145/3136014.3136031
One thing to consider, is that with JavaScript you put it in a .js file, point a HTML page at it, and that's it.
TypeScript uses a ton more than that, which would impact the amount of energy usage too, not to mention everything running the package registries and what not. Not sure if this is why the difference is bigger, as I haven't read the paper myself :)
But if you do, please do share what you find out about their methodology.
This image comes from running the different versions of the benchmark games programs. Some of the difference between languages may actually be just algorithmic differences, and also those programs are in general not representative of most of the software that runs.
That, and also because rust compiler is a very good guardrail & feedback mechanism for AI. I made 3 little tools that I use for myself without knowing how to write a single rust line myself.
I can see that a reality but I am more comfortable using Golang as the language compared to rust given its fast compile times and I have found it to be much more easier to create no-dependices/fewer-dependencies project plus even though I wouldn't consider myself a master in golang, maybe mediocre, I feel much easier playing with golang than rust.
The resource consumption b/w rust and golang would be pretty minimal to figure out actually for most use cases imho.
Stop using Electron to save massive amounts of RAM.
> Time for an AI tax on the hyperscalers.
Like the purchase price + increased cost? The thing is that these parties are sitting on billions and billions of investor money, they don't care that hardware is 400x as expensive. Which companies like nvidia have capitalized on a few years ago, they were already able to price their hardware at a 400% markup compared to pre-crypto times, and shift their focus from consumer graphics chips to datacenter compute chips, causing their revenue to go up 6x (if my interpretation of [0] is correct)
[0] https://www.macrotrends.net/stocks/charts/NVDA/nvidia/revenu...
Why tax something that the market will figure out? This is normal and things will sort themself out.
Markets only "figure things out" in a petri dish economy where:
1) There are no barriers to entry for competitors (e.g. protectionist tariffs, equal access to capital for everyone)
2) There are perfect substitutes available, so transitioning to a competitor is seamless and free (e.g. no requirement to store data in Country X, no vendor lock-in, no security compliance)
3) The industry is not a "natural monopoly" when only a handful of vendors can operate due to capital investment and national/global distribution required (see power utilities, telecoms, petrochemicals)
4) Profitability attracts competitors (won't happen because of #3), but heavy competition prevents abnormal profits from accumulating to a single player (happens because of #1, #2 and #3)
When markets don't figure things out, as is the case around the world, you get a tangled mess of market failures, government intervention and lobbying to neuter proposed interventions.
Markets are never perfect but over the course of history they are a pretty good mechanism to solve these type of problems. Not sure why we think taxing hyperscalers differently is the answer. Government usually does worse than the market when it comes to sorting it out.
My argument is not that market is perfect but that the alternatives are probably far worse, like a new tax on a specific group of companies.
Market only exists due to government.
Yes, and? That still does not mean we should add arbitrary taxes for a supply/demand issue.
We see that market is very irrational now and it can stay irrational for long enough to destroy everything we know in tech.
By the time market figures things out, you may no longer have services, and hardware that you use daily. When such amounts of stupid money are pumped into a single industry, even if all AI companies went out of business tomorrow, it's going to take years for things to go back to normal.
FWIW, I'm not advocating taxes, as I think that won't really do anything. I don't know what the solution is either.
> ...and it can stay irrational for long enough to destroy everything we know in tech.
Nah. For decades software engineers have been more expensive than the cost of buying the extra hardware needed for vastly inefficient software. There are orders of magnitude of inefficiency there. So there's a ton of slack in the world's software that can be taken up by software engineers while hardware is scarce, pushing back the date where there will really be a problem probably by decades more.
Of course software engineers will see a problem though, because they'll have to learn to to write efficient software again.
ie. "Great, but now make it work with less RAM" will be a thing again, instead of "It needs more RAM so order some as it's cheaper than your time to fix the code".
Sounds like hyperbole. Yes the world is connected yes we are seeing shortages, yes the market is imperfect and it lags but this is how things get fixed. Prices are sorted out, manufacturers make bets on long term capacity. Some will be losers, some will be winners.
My guess is that many of the current people saying "technology is over and no one will afford their own computer" might have been born after the previous shortages, so it's in reality their first shortage and they have no memories (nor interest reading about) the previous ones, that all eventually washed over, even if at those points there were also people claiming that "No one will have their own SSD in the future, because prices will always be super expensive for consumers from now on".
That's my hypothesis I spent a whole of 30 seconds thinking about anyways.
This is a different kind of shortage though. Previous ones were cyclical and caused by supply/demand mismatches or natural disasters. This one is structural. The manufacturers are actively choosing to prioritize AI because the margins are dramatically higher, and AI market has virtually unlimited money right now.
> eventually washed over
Eventually is doing a lot of heavy lifting here. Several years of constrained supply have real consequences for people and businesses. Hardware manufacturers are saying most of their capacity is already sold out to AI customers through 2026, and possibly even through 2027 and 2028, with the rest of the markets getting what's left over. This is a fundamentally different market dynamic.
> caused by supply/demand mismatches
How is that different from today? The scale might be different, but it's quite literally a "supply/demand mismatch" right now.
I don't think what we're seeing today can be described as "structural", at least because it's way too short to make such proclamations today, if it ossifies, then yeah maybe I'd agree with you, it's become structural.
> Several years of constrained supply have real consequences for people and businesses
Indeed, but lets see if it'll go as far as being "several years", the prices already stopped increasing, and supply still isn't planned to be expanded, if either of those changes you might have a point, but as of today it seems like an exaggeration.
Market is fixing it. Memory makers prioritized HBM and enterprise NAND, some, like Crucial, went out of consumer business entirely.
At the same time, the rational market is behaving rationally - they're not increasing production because they're fearing AI bubble could burst, leaving them with oversupply and expensive factories.
The market, apart from AI market, is behaving exactly as it's designed and as it should. But it doesn't mean outcome is good for everyone.
In all parts of life there are winners and losers. You cannot regulate that away.
> We see that market is very irrational now and it can stay irrational
That meme refers to speculation on stock market prices. Nobody is buying up RAM with the expectation of making speculative gains on it.
> it can stay irrational for long enough to destroy everything we know in tech
What does this even mean? I know people on the internet sometimes exaggerate, but I cannot even begin to find a more charitable meaning with this, what exactly will be "destroyed" in "tech" because of prices going up for a year or two?
Here's an easy experiment to conduct: look around the room at your home and count all the devices that have a CPU, RAM, SSD or HDD.
Then take a look at your bank statement to see what are the services you pay for monthly that also require the same hardware.
Now, imagine that these devices or services can no longer procure RAM, SSD or HDD. There's no more available supply for these components, because this is what's happening.
Would you still be able to have these devices if they all broke tomorrow? What about your hypothetical Backblaze subscription? Would you still be able to have an off-site backup?
> imagine that these devices or services can no longer procure RAM, SSD or HDD
Why would I imagine something so far out from what will realistically happen?
Again, a lot of doom and gloom over very unrealistic scenarios. Where are you even getting this from, YouTube channels?
Of course if there is no RAM or flash-storage at all available, eventually hardware will be unfeasible. But when we've experienced these sort of things before, it eventually restores to "normal" prices, and there absolutely nothing pointing to what we're experiencing now to get even worse, if anything it's already stabilized.
Valve had to delay a bunch of new products already. They also had to effectively discontinue the non-OLED Steam Deck due to the increased prices.
https://www.theverge.com/games/874196/valve-steam-machine-fr...
Yeah, which shows that Valve don't think "these devices or services can no longer procure RAM, SSD or HDD" is actually what'll happen in reality, because then they'd have to cancel the hardware fully. Instead, they're delaying it.
With a likely price increase.
Regarding OP, I don’t think they implied this will last forever, but is definitely concerning.
You don't have to look too far back in history -- look at the supply squeeze during covid, or even just during the Suez closure by the Evergiven.
Yet here we are. Markets kept growing, some companies lost during the supply crunch but hopefully we came out generally stronger. So again the doom and gloom is hard to track. Maybe we hit a couple years of different supply crunches in tech, at some point if demand sustains companies will figure it out, optimize price, manufacturing lines etc.
My laptop's 8 years old, if I can't get memory I'll just have to sweat it a little longer. Same with my NAS drive
Same with work -- I've just ordered some replacements for 13 year old servers in one office, but if it was more economical to repair them
What we're seeing is the natural conclusion of VC distortion in a market. There is so much money being pumped into AI speculatively now that it's hurting normal and sustainable businesses in other parts of the economy.
The solution might have to be mandatory rationing of some kind to avoid a situation where only a handful of AI giants are able to buy essential components. We can't just throw the rest of the economy under a bus to support the AI bubble for a few more months.
I'm working with a business right now that would like to buy some new servers for sensible, boring business reasons. It is having trouble because the prices from their normal suppliers are now extremely high - if the components are even available at all. This business has nothing to do with AI or Big Tech and yet it's at risk of being unable to continue normal operations in much the same way that a business would be affected if the phone networks were all switched off or the water supply to its office was cut. We regulate those industries because their continued reasonable operation is essential to make sure everyone else can continue to operate reasonably as well.
I'm seeing the same thing. I was consulting a group of people in my city that wanted to digitize massive load of old VHS tapes. No AI, no crazy tech, just standard, boring storage+network infrastructure.
I'm looking at the procurement sheet that I made for them a year ago. Half of the items are no longer available, while the other half became so expensive that we'd probably build 10 of such labs with these costs a year ago.
I'm also looking at my home NAS right now - I pray not even a plastic clip breaks inside, because I'd have to shut it down.
While these are still likely the first things that you'd think of being affected, I'm sure the effects are rippling through essentially every industry that utilizes these components in their supply chain. Which is probably - every industry nowadays?
I think that’s a massive stretch. What we are seeing is a new frontier in tech that nobody knows where it will land yet. Hyperscalers see a future where if they don’t build now that they might be left behind.
Absolutely VC money is flowing around but I think it’s unclear where the cards fall yet.
Not sure what you would regulate here. I hate the tripe that America and China are at war but I do think it’s not a great decision to stop the current work the west is doing as China is pushing full steam ahead.
I wish this comment can be on the absolute top of this page. This really is one of my frustrations with the AI bubble.
Fwiw, the days of creating an good ol' reliable hosting provider/Vps provider are over. I looked extensively into it one time out of curiosity but this would be one of the worst times in history to do that.
We would be sort of stuck with the options that we have right now and more and more shops in Lowend are even shutting down or raising prices with the sheer ram crisis and even HDD and storage crisis now.
A provider in LET had a post which said, "what should we providers do to deal with the ram shortage/ram prices"
These providers gave competition/had different unique features too to have chosen them but they were also incredibly price sensitive and the AI bubble blew the sensitivity by raising the prices almost 5 times or more. This would impact real businesses.
Thank you for creating this comment. I hope more people can read this. I genuinely just want this bubble to burst asap so that we can see a sense of rationality back within the market/the market functioning as expected without the immense irrationality/unpredictability of future.
another point is this, from my hosting provider idea, I shut it down. Why? because it literally makes 0 sense to start now, its postponed indefinitely untill the bubble bursts/ram prices are decreased.
How many other projects might be going through something similar. Gck1's comment next to mine also gives an example of a project whose value of cost increased 10 times.
How many of such projects would simply be unable to be built because of the ram inflation can't be underestimated imo.
and forget people who wish to game and many other things too. Basic comodities in the previous year or two feel like luxury now. All because of AI. It's insane.
They didn’t the last two memory crunches. Litigative action figured it out first.
Because this perfect version of capitalism you think exists, doesn't.
We live in a world with markets dominated by cartels of tech companies who don't play by the rules. Every other industry that impacts society in a negative way typically pays some sort of specialized tax to offset that, I don't know why these tech oligarchs shouldn't have too. It's wild how people just want to let them do whatever they want.
Everyone says we need to deregulate tech, and certain industries to get ahead of China.. Isn't it funny how their largely government controlled economy (to a degree) is annihilating the west on all fronts economically. We need far more regulation.
China will defeat the West solely because it regulates its billionaires, not the other way around like we have it in the West. And I hope so, the world is rooting for you China.
Way to put words in people mouths. Markets are imperfect but I do believe on average they are one of the better tools to solve supply and demand issues.
I don’t know who will come out winners but I do agree that China did well taking the playbook from Singapore and navigating their country through incredible amounts of growth. They are still facing depressing housing prices and deflation in other parts of the economy.
There are absolutely areas where markets breakdown, thinking problems where impacts are on longer horizons but for simple supply and demand like what we are seeing today, things will sort out in a couple years.
Reads a bit like the Paperclip Maximizer appearing way ahead of schedule? Implemented not as AI, but as emergent behavior in the ways of the financial class (that happens to be about AI, singularity and all that).
These changes are effective April 1st for existing and new customers. The price increase ratios are also different across product lines.
* Cloud (VMs): 38%
* Bare metal: 15%
* Memory add-on for bare metal: 575% (effective immediately)
It feels like memory add-on is intentionally set high to discourage customers from adding more memory.
AX102 (128 GB RAM) costs €124, AX162 (256 GB RAM) costs €244, but the 128 GB memory add-on alone costs €264. If we ignore the setup fee, it’s more cost-effective to provision additional servers instead of adding RAM to bare metal instances.
Here's the link to cloud and bare metal pricing changes: https://docs.hetzner.com/general/infrastructure-and-availabi...
> * Memory add-on for bare metal: 575% (effective immediately)
> It feels like memory add-on is intentionally set very high to discourage customers from adding more memory.
Memory prices are so stupid now that 575% is pretty close to their actual costs.
https://pcpartpicker.com/trends/price/memory/
DDR5-6000 2x32GB: ~$200 -> ~$1000
Have you seen the price of RAM recently?
AFAIK, it's been stabilizing lately at the current price, so at least it's not increasing anymore: https://pcpartpicker.com/trends/price/memory/
By the same time next year the prices likely gone down, although maybe not to the pre-increase, but surely much lower than currently. Putting it in my calendar to revisit this comment in a year :)
Stabilized at 5x (or more), a change that occurred over like 3mo.
Grocery prices have also stabilized but I’m still paying too much ha
Well, if all the doomers and gloomers were correct that this is the end of hardware at home, we'd see the price continue to increase, and suppliers trying to ramp up production, even if it'd take long time.
The fact that it stabilized (at whatever price) and that suppliers aren't even thinking about ramping up production, should tell people that the doomers and gloomers were yet again over-reacting to things they don't fully understand themselves.
> Grocery prices have also stabilized but I’m still paying too much
I think that's a local problem, if you happen to live in a country that's trying to move over to isolationism rather than globalism as of late. In other modern countries the prices are also increasing, but at least following inflation somewhat so the increase doesn't seem as bad for us. Maybe at least yet? Who knows.
> Well, if all the doomers and gloomers were correct that this is the end of hardware at home, we'd see the price continue to increase, and suppliers trying to ramp up production, even if it'd take long time.
Ramping up production takes months and paying back the price to ramp up production takes years. Manufacturers have started investing in more production capacity but it'll take a while before supply can be sold off.
Based on interviews with industry professionals, I believe the forecast is that RAM prices will start going down again between August and the end of next year. Until then, prices will climb as stock depletes and RAM production is capped.
> Manufacturers have started investing in more production capacity
Where are you getting this from? Because that's not what I've seen, if anything the industry seems to lowering the production capacity, not increasing it.
And even if it takes years, if they thought it was a sustainable growth in demand, they'd at least be moving in that direction which again, doesn't seem to be happening.
> I believe the forecast is that RAM prices will start going down again between August and the end of next year. Until then, prices will climb as stock depletes and RAM production is capped.
You're already wrong with this today, prices stopped climbing already and have been stabilizing at the current prices... https://pcpartpicker.com/trends/price/memory/
Smart money says by Jan 2026 ram prices will not be anywhere near where they were 6mo ago, but we’ll see I suppose.
> * Memory add-on for bare metal: 575% (effective immediately)
I don’t see this anywhere, source?
I used Hetzner's pricing calculator.
https://www.hetzner.com/dedicated-rootserver/ax162-r/configu...
Before today, we used to be able to order an AX162-R for €207 and add 128 GB of RAM for €46. Starting today, the same calculator provides €207 for an AX162-R (*) and €264 for the 128 GB RAM add-on. Sadly, HN doesn't let me upload screenshots.
(*) The price change for AX162-R machines is effective starting April 1st.
Yeah, not sure where they're getting those from.
From the Robot UI, I tried ordering a new EX44 or EX63:
- EX63 comes with 64 GB DDR5 by default, can be upgraded to 192 GB DDR5 ECC for added €42.35
- EX44 comes with 64 GB by default, can be upgraded to 128 GB DDR4 Non-ECC for added € 16.94 max. per month
> These changes are effective April 1st for existing and new customers.
Checking today doesn't really indicate anything.
It's worth noting that the hardware price of RAM is up at least 550% yoy, so this was always going to happen as soon as their existing contracts had to be renewed
> > These changes are effective April 1st for existing and new customers.
I thought the "effective immediately" mean that April 1st threshold wasn't for the memory...
This was already discussed, but that post got dumped onto page 5 after just a couple of hours for some reason.
https://news.ycombinator.com/item?id=47120145
I really love that their notification email includes applicable price change for my specific servers.
The worst counter example of this was Mercedes sending me an email saying "the terms and conditions have been updated, please read them at this link". It linked to the 52 page document I was supposed to read through in its entirety and manually diff against previous! Good thing they started adding a change log in the emails after some customer push back.
My ISP sends me an SMS telling me there's work being done in my area. I have 3 different accounts with them in different citites. No idea which one they are talking about at any given time.
Hm the pricing increase stresses me out out less than the server shortages. My impulse reaction is to buy a few cheap cloud VPS instances even though I don't need them right now... Anyone have any wisdom to encourage/discourage this?
I just bought a Raspberry Pi 4 1 GB memory with aluminum case, aluminum NVME adapter, and a 64 GB SSD for about 80 euros. With microsd it’s even cheaper. 4 GB RAM would be about 120 euros.
The 1 GB RAM replaces one Forgejo runner that was in Hetzner. With €5 per month, I will earn this investment back in less than two years. After the price increase, this period will only shorten!
I also wrote about this at https://huijzer.xyz/posts/148/raspberry-pi-as-forgejo-runner
Still cheaper than US cloud computing.
In EU there are: Hetzner, OVH and Seeweb.
https://european-alternatives.eu/
Also Contabo
Also Scaleway
I like Scaleway a lot too.
Doesn't seem to apply to older/deprecated gen instances. I've got a CX22 there for personal screw-around projects and it's the same £3.95/mo (pre-VAT) afaict. So maybe not much help to folks ordering new or running on the current gen as the older kit isn't something you can order now, but a small boon for us laggards.
For some reason I didn't get an email from them about this, even though one of my VPSs is in Helsinki.
Anyway, let's all please pretend that Hetzner is now way overpriced if anyone asks about it. :P
I almost didn't see their email because it's sent by "notification" (notification@hetzner.com, no name set). Title is "Update on our pricing".
Anyway my increase is: EX44 (HEL1) € 44.76 -> € 50.76
Not pleased especially as the reason for the increase for existing customers is a nebulous "The costs to operate our infrastructure has increased dramatically."
Why?
Maybe I have email notifications off.
I think it's RAM and server prices globally shooting up. Extreme RAM shortage and increasing hoarding of all types of hardware supplies.
I am confused why the announcement page says CCX33 in USA "Old price" is €59.49 but their main pricing page shows €50.49 for CCX33 in USA
Announcement page: http://docs.hetzner.com/de/general/infrastructure-and-availa...
Pricing page: https://www.hetzner.com/cloud/
I assume that's with/without tax. Those German prices would be inclusive of their 19% VAT.
Western memory manufacturers decided to chase the AI bubble, abandoning the consumer and low-requirement markets entirely.
Chinese manufacturers are now capturing that entire segment with full vertical integration. When this bubble stabilizes, because it will (it's not going to grow to infinite), Western companies won't recapture those markets.
They've already ceded competitive advantage for the next decade. This is a structural shift, not a cyclical shortage.
It's another step in the transformation of Western industry that began in the '80s: the shift from real economy and human-centric production to financialized operations.
It's not just Hetzner cloud; got an email about increase prices on my dedicated server.
Good. This means the market is healthy.
Hopefully this also means new providers appear in Europe, to handle the increase in demand.
If you just want an app server pick up an hp elitedesk off ebay and a ups and run it on your home inet connection.
I've just been looking in to this as I've got quite a lot of older hardware that'll be fine for running some websites lying around.
My ISP has a static IP option for £5/month, but I reckon I can save £30/month+ on server costs even before any rises.
Ofc it does mean I have to do my own sysadmining, but a combination of my general knowledge + an LLM should make that relatively easy.
Watch out for the energy usage. What's electric now, 27p/kWh?
Note: your ISP might cut your account, and this might violate your home insurance contract — depending on jurisdiction.
At least where I live there’s a stupid amount of red tape for these things.
Your ISP will cut your account when you saturate the upstream pipe 24/7 for weeks on end... which will only happen if you host video.
And your home insurance will not know/care if you're operating a desktop-sized computer or even a single server (it is perfectly fine and expected a developer might bring an actual server home for troubleshooting). Home insurance only cares if you're running dozens of them.
It depends. If you're running a business from home is where my local insurance draws the line and you need additional professional insurance.
You should be able to do this hassle free, and you probably can get away with it, but you may find yourself in a grey area later.
It's just one of many types of red tape that stiffles innovation.
IF you just want a Pizza, pour some tomato ketchup on sliced bread.
If you just want to pilot a 747, drive your car really fast at a skate ramp.
A 1000 calorie pizza is often overkill for a meal
A 747 is overkill for a fetching some groceries..
Somewhat weirdly I’m very happy about this price increase as a customer. The messaging is clear and completely understandable. Well done.
This excuse "we need to raise prices because we have more demand" is BS. They should be truthful and say "we can increase prices and people will pay it because they want to be EU based"
To be honest for anything more serious than a personal Minecraft server hetzner has been beaten by ovh for ages (on bandwidth - you get all you can eat data limited by speed from ovh - for example 500mbit, instead of 20tb from hetzner).
For this reason hetzner is always a "backup DC" in my eyes and never the primary.
Also I heard they are extremely sensitive regarding abuse allegations so don't even think of hosting something someone may not like seeing...
They get a lot of hype, but there are many competitors worth looking at.
They're using Arbor, they were cheap for that exact reason.
Now that people don't care about Anti DDoS - this happens.
In the past everyone was leaving Hetzner for the OVH/Voxility due to terrible latency and nonexistent protection.
This comes after OVH sent emails with really spicy increases too. Like north of 50
I got no such email, was this for VPS or dedicated servers?
Not sure which products but Reddit etc was in uproar
https://www.reddit.com/r/OVHcloud/comments/1ra5jzg/ovh_doubl...
Hetzner had the best prices out of any cloud I’ve used. Sad to see that they are raising prices, but was due to happen.
Still a fraction of the cost of most other providers, and wouldn't shock me if we see the others all doing something similar.
companies that haven't turned a profit are outbidding the rest of the economy for hardware. that's not a supply shortage, it's a subsidy funded by venture capital.
36% as per the linked post, 38% was a typo.
On one hand this is not good but predictable. I'm on longer-term commitments with OVH, so it will be interesting to see how they follow. I'm still keeping Hetzner on my shopping list, even with the increase the bare-metal offerings are within my budget, and now that prices have increases they should be stable for a while (also import for budget management).
I moved from paying 24.50 a month to 25.39 a month for my little VPS plus storagebox.
CPX31 Cloud Server (Germany): €13.10 → €13.99/month (+€0.89, ~+6.8%) BX21 Storage Box: Unchanged Primary IPv4: Stays at €0.50/month
I haven't received this email, and I have one x64 server that costs around 4 EUR/mo, and an ARM server that costs about 6 EUR/mo. I wonder if I'll still be affected by the price increase.
They've only ever increased the ipv4 prices for already existing customers before if I am not mistaken. This is quite big.
EDIT: It's not a huge increase for dedicated servers. I already can't find anything comparable for more than the increased end price.
> AX51 (FSN1) € 63.10 € 64.99
> AX101 (FSN1) € 107.10 € 110.31
> They've only ever increased the ipv4 prices for already existing customers before if I am not mistaken
No, that's not true, they've done increases before, at least for VPSes only, I think that was 1 or 2 years ago or so?
You might be right. Sorry. I should have said they haven't increased prices for existing dedicated servers since my direct experince is only there. Actually until about 3-4 years ago when the whole world went to shit, using a server for a year or two then upgrading to a better server for cheaper, was the norm. In that environment, you would naturally not have price increases.
I also don't think you're right that it never happened for the dedicated servers :) I'm only using Hetzner for dedicated servers, and found an email from 2022 where they mention price updates:
> Unfortunately, we are forced to increase the prices on these Server Auction models [...] old price 37.60 Euro -> 59.29 Euro, comes into effect 2022-03-03
Citing raising energy prices at that time.
Probably not for existing customers (their existing servers). I don't recall anything other than the IPv4 related increase in the past. I might be wrong of course as I've demonstrated already.
> Probably not for existing customers
Yes, in 2022 I was an existing customers, and my server increased in price then, the server affected at that point went from 37.60 Euro to 59.29 Euro. Today that same server went from 65.22 to 67.18, so there is even more price increases seemingly between today and 2022 but I'm not finding exactly when that was.
My CCX13 (dedicated cores) went from 15€ to 20€ now. Looking at Netcup as alternative, more cores and more RAM for 12€ - anybody has experience with their root (kvm'ed) servers?
It's OK until you get into support troubles. I would say Scaleway/OVH might be better contender (but they are french hehehe).
This is likely just the first wave. If this component hoarding by AI continues, and it likely will, at some point, it will be just OpenAI and Anthropic who can afford to have compute.
This has affected SSDs first, then RAM, then HDD and it doesn't look like even HDD manufacturers are going to increase production. So unless groups of people suddenly learn how to manufacture all of this hardware and open factories quickly, it's going to be a very fun next few years.
People have been predicting SaaS will die for all the wrong reasons. It's not that anyone can ship a SaaS clone by prompting an AI, it's that nobody is going to have access to the hardware required.
This mirrors the increased costs of people who already space + power in a DC, and want to buy new machines to fill their racks. Everybody is being hit.
With the recent price spikes in memory and storage, this was just a matter of time.
Ah yes, the abundance of AI that keeps on giving.
Wow. That sucks. hcloud was great for ages and highly competitively priced.
Vultr may be a good alternative. If you want to search VPS prices across the 6 major clouds (gcloud, aws-cli, hcloud, az, doctl, and vultr-cli) I made a wrapper TUI that lets you search, sort, and rent VPS.
See it here: https://tui.bluedot.ink
I would be very surprised if all hosting providers didn’t increase their prices eventually.
This is it. Hetzner has always been very price competitive in its existence. Given the private ownership, I din‘t expect this to be a sudden outburst of greed, but to actually reflect rising costs.
If a provider has higher margins, they may choose to eat some of the cost. But I would not expect that to be the case across the board
> Vultr may be a good alternative
I feel like a huge selling point of Hetzner is that they're based in Europe, and they're themselves citing that as the reason for a huge uptick in sales and new users. In that context, I don't Vultr is a realistic alternative.
OK, I never thought of it like that. It was always a price thing. For a while Vultr and Hetzner were much better value per unit.
What's behind the European push?
> What's behind the European push?
Obviously the US pushing absolutely everyone away and making EU and Europe the new enemy, so now we here want to reciprocate that and feel the need to move away from US infrastructure ASAP.
Personally I've been on a personal quest to minimize my usage of US-based services for many years already, but right now it's even part of the mainstream conversations, so seems to be ramping up, finally.
Is this voting with dollars (euros) due to views, or is there a regulatory reason to avoid US providers in Europe?
For clients, I just do what they wish to do, and a bunch of them want to move to European infrastructure because they've seen what can happen when you rely on US infrastructure today, and don't want it to happen to them. Only one so far cited regulatory reasons, and I think they were misinformed, but helped them out anyways with it.
Personally I do it because it's better aligned with what kind of future I want, and not wanting to support hyper-captalism environments anymore.
Cloud Act directly conflicts with GDPR. To really rub salt in the wound Trump overtly threatening to invade the EU (Greenland) basically turned the whole of Europe off seeing the US as a reliable ally. I don't think Americans have caught up to how much damage he has done to the image of the US amongst allies. They seem blissfully unaware of what's happening. Of course there are plenty of astute Americans who are aware but not the public at large.
This will be as a shockwave in web hosting industry, the same as it was with electricity price. There is nowhere to run. Everyone will increase their prices, unless hardware crysis ends up.
OVH increased prices by even more. So no reason to move.
Even my more then 11 years old server increases by 80 Eurocent! Dare you!
How much is the cost for Storage Boxes increasing?
I don't see them listed on the announcement page (BX* products), so I'm guessing storage boxes prices will stay the same.
Surely that means that as soon as prices of ram drop, Hetzner will also drop the prices, right? RIGHT?
Hezner reducing prices is not unprecedented. I think RAM chips getting cheaper is a less likely event than Hetzner responding with dropping prices
At best they would freeze prices for a few years which would be a real term decrease
[dupe] Discussion on source: https://news.ycombinator.com/item?id=47120145
My increases were around 4%
I recommend Netcup as a solid EU budget alternative to Hetzner, zero complaints from me.
This is probably the best alternative provider for individuals that I could find, unless you're orchestrating a fleet of servers or something. Personally I'll wait out my next billing cycle with hetzner, as I expect other hosts to follow shortly.
They are solid and cheaper, but they don't offer the same level of control plane and API access as Hetzner that is really helpful when managing a larger number of servers.
Their ARM64 boxes are fantastic, but sold out at the moment.
"Edit: It's 36% ! Can't edit the title typo of 38%"
Ouch. OVH are also going to increase their prices.
... more customers so they must increase prices? This seems backwards from how scale usually works.
It is, but more customers at a time of historically high component prices will do it. If you set your costs assuming every user's hardware is $1, and your customer base doubles when the hardware is $2, you're going to have to raise prices for everybody
The next set of hardware purchases will cost more than their last set of hardware purchases, and that's going to outweigh any labour economies of scale given just how many hardware components are in shortage this year.
If their growth had been in their projections in say 2024, they might have just been able to skip a round of hardware purchases, but the combination of growth meaning they must expand their hardware and hardware costs made this inevitable.
Can anybody predict this craze? The classical memory manufacturers are not yet adding additional manufacturing capacity. They learned this hard way in the past. That means, the demand is here to stay for years without typical bubble burst. Is this a point where Chinese companies will rise worldwide?
The massive DC overbuild matches demand, prices normalise somewhat in 3-5 years.
The massive DC overbuild does not match demand, prices tank in 3-5 years.
Third possibility: some approach like Taalas renders the current storyline meaningless. Would put 3 in 10 odds of this happening but I'd looove to see it.
Fourth: entire planet gets profoundly sick of emdashes, we all move back into caves and live in eternal gratitude of the moment humanity woke up to how little all of this really matters.
Hard to predict. If the bubble pops (NVIDIA and "circular economy", massive FAANG datacenter expansion plans, huge LLM training budgets) the markets will once again be flooded with components.
But, the shortages may very well continue into 2027, leading to some manufacturers going out of business and yet another massive redistribution of wealth.
I just hope the whole thing comes crashing down and we can buy GPUs and RAM again.
I mean I might not have a job in that economy, and my pension might be screwed but I'll have 192GB ram so I should be fine.
Silver lining: can you imagine how dirt cheap RAM will be after that bubble has popped? Oh my...
It won't. Demand is being pushed forward. That means that longer this situation take longer it will take for prices to recover to same levels.
RAM producers aren't adding more capacity on the non-HBM side of things, so we shouldn't see a dramatic drop in pricing if AI HBM memory demand drops.
No manufacturer is increasing supply though. RAM, SSD, HDD - they just reallocated their existing supply to AI.
Your home systems can slot in HBM? Doubt that.
This is a simplistic view of why the prices are the way they are.
... and still remain far too cost-effective. Frankly this says more about the rest of the industry than for hetzner
BuyFromEU is the funniest subreddit there is right now. Unintentionally but still entertaining. EU has managed to paint itself into unenviable corner. I can't buy from EU even thought I want to because for physical goods - cross country shipping costs are prohibitive and for digital - they are either subpar, more expensive or both.
Try this as experiment - try to buy something like precision dowel pins from Poland or DOLD Mechatronik with shipping to Greece, Bulgaria or Romania vs the same thing from Aliexpress or Temu. Chinese costs are cheaper even if they have to fly here.
QC / Cheap Shipping / TEMU or AliX Pricing
Pick 2.
Not to mention that from July 1, 2026, the EU is abolishing the €150 duty-free threshold for non-EU shipments. This is specifically targeted at the flood of packages from marketplaces like Temu and Shein.
From July there will be a flat customs duty of €3 for small consignments. This fee applies per category. If your package contains items from different product groups (e.g., a shirt and a cable), you might pay the fee multiple times.
The Goal: To create fair competition for European retailers who can't compete with subsidized shipping and tax loopholes from massive non-EU sellers.
This will obviously have a knock-on effect for larger shipped items which are presumably subsidised at the bottom line by these parcels of fast-fashion and eWaste.
As someone that frequently buys low-cost second hand electronics from Japan, I am a little frustrated about the €3 per-category customs duty. That means a €80 package of various old game cartridges, retro handhelds, digital watches and collectables will now have another €12 to €24 on top of the 21% VAT and €6 handling fee. For an €80 package I am now looking at €15 for shipping and €34 to €46 in import cost. That kills a fun hobby.
The reverse is true too. The shipping costs from US -> EU are prohibitively high, oftentimes making ordering stuff from the US a no go. I think only Amazon (for certain products at least) charges somewhat reasonable shipping costs.
China has an enormous leg wrt shipping.
>Try this as experiment - try to buy something like precision dowel pins from Poland or DOLD Mechatronik with shipping to Greece, Bulgaria or Romania vs the same thing from Aliexpress or Temu. Chinese costs are cheaper even if they have to fly here.
This is an awful experiment. Only consumers care about delivery costs on deliveries like these, and what you're looking at are explicitly not goods aimed at consumers.
Okay. Then buy pizza oven from Italy and see how shipping costs are 60% of the price of the oven.
Anyway - you seems to misunderstand. If transporting something from Shenzhen to Franfurt is cheaper than transporting the same thing from Krakow to Thessaloniki - means that EU has fucked up royally in its main mission - to facilitate movement of goods. WE have ungodly patch of local carriers and courier companies and a lot of friction in every kind of intra eu goods movement.
> Then buy pizza oven from Italy and see how shipping costs are 60% of the price of the oven
Is it even physically possible to deliver at a significantly lower cost? Pizza ovens are both very large and very heavy, you can't fit many of them in a vehicle. They're also tricky to load and unload.
> If transporting something from Shenzhen to Franfurt is cheaper than transporting the same thing from Krakow to Thessaloniki - means that EU has fucked up
Ummmm. No.
It means the United Nations Universal Post Union international treaties which effectively provide China with subsidised postage TO THE WORLD (as China is a "developing country") needs urgently updating....... Some of the postage you pay to send parcels within the boarders of your own country is used to subsidise crap posted from China.
https://en.wikipedia.org/wiki/Universal_Postal_Union