Comment by dannyw
5 hours ago
NVIDIA is one of the most open labs.
They even share many of their pre-training and even post-training datasets for Nemotron on HuggingFace; for example: https://huggingface.co/datasets/nvidia/Nemotron-Post-Trainin...
Which other lab shares this?
Yes, there is no question NVIDIA wants to lock you into CUDA and their hardware. But also, they’ve consistently demonstrated the most openness when it comes to model training, datasets, and research; even before the LLM era (e.g. StyleGAN).
There’s also modelscope.cn (china’s huggingface) which is worth checking out. I would not be surprised if one day, we have to use China VPNs to download open weight models.
> NVIDIA is one of the most open labs.
Of course they are. They're commoditizing their complement.
I want to own the hardware, not play around in an nvidia fiefdom full of nvidia rules.
I'll expand on the economics somewhat. My intuition is that you can like a moat left of you in the supply chain, but dislike moats to the right of you. (I'm using left and right as I picture this horizontally drawn. It is usually called vertical integration by economists.) But free competition in your market is worst of all. Free competition right outside your moat is pretty sweet, and that's were parent is commenting on.
Nvidia probably likes that ASML is a monopolist (it's called a monopsony). The price is high and Nvidia can't scale as hard as they want to (more general: capital intensive market). This monopsony makes sure that other chip companies can't rapidly scale up and try to beat Nvidia. That there only ever was one other GPU firm (I'm prehistoric; once there were more) and they bungled it on software, is pretty sweet for Nvidia.
On the side of their customers. It would be best for Nvidia if there is free competition for the outputs generated from there GPUs. This maximizes consumer surplus and thus demand. Maximum demand for tokens, is maximal demand feeded in their monopoly. If there is a monopoly right from you, demand is curtailed, and your value is limited.
And that exactly is why you see interest from token generators for chips. Bridge that moat and gain a larger value surplus. Both NVidia and, say, China actively undercutting the token-supplier value chain is quite interesting to watch. It's like the Opium wars with us as somewhat happy customers.
In this same vein, why isn't ASML raising thousands of billions for building their own (subsidiary) chip foundries, while raising prices and starving the market (a little) for their machines.
While nvidia 's drivers are closed source there's enough interest in running LLM's that you are not locked in using nvidia. Strix Halo chips have been around before dgx spark came out and deliver very similar performance.
You do own the hardware. Nvidia GPUs have FOSS driver alternatives (Nouveau, NOVA) and have supported OpenCL on the proprietary drivers since 2009.
If you don't want to use CUDA, they expose the PTX bindings to write your own CUDA alternative too: https://docs.nvidia.com/cuda/parallel-thread-execution/index...
When last looked at it, NVIDIA was not supporting OpenCL beyond 1.0.
Also, when running OpenCL, NVIDIA hardware disables multiple DMA engines, and allows only one memory transfer at a time to prevent OpenCL running as fast as CUDA.
Did NVIDIA finally allow open source drivers to access all parts and features of the card to allow feature parity? Last time I checked they were considering a plan for planning a solution to that.
2 replies →
I wish OpenCL was usable today like your implying...
1 reply →
nothing stopping you from doing that
Are you joking? They just bought HF for $13bill and didn’t break a sweat. They are dominant and spreading.
Can't buy a decent modern Intel gpu to run AI on because of nividias "donation"
1 reply →
Nvidia has closed drivers, which makes them light years behind AMD in trust.
Also if we were just discussing labs, Ai2 opens ~everything with dramatically less resources than Nvidia.
> NVIDIA is one of the most open labs.
Nvidia’s history with linux shows the opposite. And as a user running models on a linux/AMD stack, this information does not fill me with hope.
_why_ have people started to call companies “labs”?!
Because making new AI models is a research activity, so the group within Nvidia (or other companies) that does that activity is called a research lab, or just "lab" for short. I don't think anyone is saying that all of Nvidia is a lab (that's just shorthand I guess).
Their (not especially great, compared to Chinese ones) models being open weight doesn't even come close to outweigh the effect of CUDA & Co being proprietary and closed.
"The proprietary shovel seller has some excellent tutorials on how to dig gold. Nobody else has such good step by step guides. Therefore them buying a shovel-agnostic tutorials and techniques method (that has a lot of info on using other shovels effectively) is justified."
I'm with Torvalds here, fuck you Nvidia.
They are sharing stuff that makes you consume their stuff. Their position with big ai companies is always vulnerable (asics/direct tsmc relations)
> They are sharing stuff that makes you consume their stuff
No shit Sherlock. Name one company that shared their code to make you NOT to consume their stuff?
Not really just to counter Anthropic
> NVIDIA is one of the most open labs.
Dude, where's the src for GPU drivers and the firmware blobs?
Sorry, you can't say it is one of the most open labs without qualifying a proper response to the question above.
"Labs" here specifically means "AI/LLM model lab", which that part of Nvidia is unquestionably one of the most open.
Nvidia is also one of the most closed hardware developers around. Two things can be true.
Then this whole comment thread was a diversion though, as it was specifically about the hardware parts and not model training.
NVIDIA is a lot more than a “lab”.
lol “Everything you said is correct but you’re wrong because I’m thinking about something other than what you were talking about”
CUDA is proprietary for a pretty understandable reason. There's no good way for Nvidia to standardize it.
They supported OpenCL when Khronos floated the idea of a GPGPU standard to manufacturers, but OEMs didn't want to design scalable hardware or sponsor the software.