←back to thread

Getting DeepSeek-OCR working on an Nvidia Spark via brute force with Claude Code

(simonwillison.net)
200 points simonw | 1 comments | 20 Oct 25 17:24 UTC | HN request time: 0.2s | source
Show context
cat_plus_plus ◴[21 Oct 25 23:11 UTC] No.45662944[source]▶
No idea why Nvidia has such crusty torch prebuilds on their own hardware. Just finished installing unsloth on a Thor box for some finnetuning, it's a lengthy build marathon, thankfully aided by Grok giving commands/environment variables for the most part (one finishing touch is to install latest CUDA from nvidia website and then replace compiler executables in triton package with newer ones from CUDA).
replies(2): >>45663192 #>>45665204 #
1. ur-whale ◴[22 Oct 25 05:31 UTC] No.45665204[source]▶
> No idea why Nvidia has such crusty torch prebuilds on their own hardware.

NVidia is a hardware company at heart.

They do create and sell amazing hardware.

But like almost all hardware makers I know, they totally suck at software. Their ecosystem is an effing nightmare (drivers, compilers, etc...). It's a pure culture issue, where:

   a) software always comes as an afterthought

   b) the folks in charge of engineering are largely HW folks. They like like HW engineers and the resulting software stack looks exactly like a piece of silicon: opaque, static, inflexible and most important of all, never designed to be understood/looked at/reworked.
I suspect the reason why all their software is closed-source is not for commercial reasons, they're just ashamed as a company to show the world how shitty their SWE skills are.