Exact Gradients Through a BRep Kernel

(muni.works)

27 points | by ecto 3 days ago ago

10 comments

  • JoeDohn 2 hours ago

    My first thought when I read (skimmed mostly) the paper is PGO (Profile guide optimization), except instead of optimizing machine code based on runtime profiling, this would optimize geometry based on physical or engineering objectives ?

  • YuechenLi 2 hours ago

    So, the Claude English to human English translation:

    Ordinarily, if you want to adjust the parameters of a CAD part, say, a bracket, to meet a constraint like a particular mass or center of mass, the approach is to change every parameter a little, rebuild the geometry to see what physical properties changes, and repeat until you get the result you wanted.

    This is of course a very expensive operation because you have to iterate through every parameter on a part every single time for any optimization pass.

    So VCAD's approach is to essentially record how each of the mesh verticies connect to each other after the BRep tesselates, then calculate what part of the mesh are actually affected when a parameter change, and then use that to guide their optimizer to get the physical result they wanted, so they only need to change the relevant parameters they found that affects properties instead of doing the computational equivalent of banging your head against the keyboard and hoping something works.

    • bschwindHN 2 hours ago

      Thank you! It's a shame there are so many LLMisms in this paper, it really cheapens the whole thing and makes it harder to understand than it needs to be.

      Write papers and articles in your own words, people!

      • ecto an hour ago

        Frankly, I have not the time. I have a son, a business, and research to complete. Let people live. I don't care if the words cheapen it for you - I present my ideas for the world to enjoy.

        • programjames 18 minutes ago

          I think it is pretty well done and far above the quality of most papers. I also think the prose is not up to the same standard you applied for the places where you made choices (e.g. presentation, layout, communicating key ideas). But I totally understand where you are coming from. It takes a ton of work to apply theory of mind to figure out how to say things intuitively for your audience (the main gap of LLM prose here), probably 10x the work of carefully choosing what ideas and in what order to communicate them.

        • iamsam123 43 minutes ago

          Presenting your ideas is one of the most important parts of research…

          • ecto 36 minutes ago

            The ideas are presented how I intend, after months of work.

        • LtdJorge an hour ago

          Does that mean you don’t care about the time of people reading your research, if it makes it more difficult to read?

          • ecto 38 minutes ago

            Your reading comprehension is none of my concern.