"The goal for this year is [...] 1000 developers on Discord."
Am I understanding this correctly, they want 1000 people on their Tenstorrent Community Discord or something? Are those the kind of metrics we're working with nowadays? If so, that's bleak
If they want to onboard developers they need to seriously question their approach to software. Clearly, they've not been building the kind of abstractions developers want.
The chips don't strike me as particularly power efficient (especially at idle which matters for developers), or cost efficient (by any significant margin). And on programmability, they're losing big to everyone else. They need to win in *at least* one area for people to be interested.
I wish they focused on making a GoogleColab alternative where we could use Tenstorrent hardware.
They're pushing towards devs but cite not a dearth of accessible devices.
I have neither an Nvidia GPU nor a Google TPU but guess what: I'm fairly competent working with both because they've been made accessible.
"The goal for this year is [...] 1000 developers on Discord."
Am I understanding this correctly, they want 1000 people on their Tenstorrent Community Discord or something? Are those the kind of metrics we're working with nowadays? If so, that's bleak
please join the discord :)
Gimme free hardware first ;).
If they want to onboard developers they need to seriously question their approach to software. Clearly, they've not been building the kind of abstractions developers want.
The chips don't strike me as particularly power efficient (especially at idle which matters for developers), or cost efficient (by any significant margin). And on programmability, they're losing big to everyone else. They need to win in *at least* one area for people to be interested.
Praying they succeed.
Edit: They are winning on scaling/multi-chip.
I wish they focused on making a GoogleColab alternative where we could use Tenstorrent hardware.
They're pushing towards devs but cite not a dearth of accessible devices. I have neither an Nvidia GPU nor a Google TPU but guess what: I'm fairly competent working with both because they've been made accessible.