6 ms·

I heavily disagree. Nvidia has a strong interest in open source AI being a strong contender to proprietary models. They don't want a single AI company to win,

I heavily disagree. Nvidia has a strong interest in open source AI being a strong contender to proprietary models. They don't want a single AI company to win, as that company would then have enormous leverage over Nvidia, and could perhaps even erode the Cuda moat with their own chip spending. They want the AI market to be hypercompetitive, with AI companies focusing on competing amongst themselves instead of with Nvidia on one side, and many self-hosters / small inference providers buying Nvidia GPUs for large open models on the other. It's in Nvidia's business interest to be a good steward of Huggingface, while being a good steward of Github is at best incidental to the interests of Microsoft.

fjcururuvy7 2mo ago
HN failing to understand embrace extend extinguish yet again I see.
- hanselot 2mo ago
  EEE is impossible already. They can't close pandoras box now. Why is it alarming for a shovel salesman to also sell plants?
  - AnthonyMouse 2mo ago
    Context: Liz Lemon has been buying two products together because one of them creates a need for the other. Jack Donaghy: The only thing I will be discussing with the House Subcommittee on Baseball, Quiz Shows, Terrorism, and Media is vertical integration. Liz Lemon: What's vertical integration? Jack Donaghy: Imagine that your favorite corn chip manufacturer also owned the number one diarrhea medication. Liz Lemon: That'd be great, 'cause then they could put a little sample of the medicine in each bag. Jack Donaghy: Keep thinking. Liz Lemon: [beat] Except then they might be tempted to make the corn chips give you... Jack Donaghy: Vertical integration. https://www.youtube.com/watch?v=ZZ7oht6TD9c https://www.youtube.com/watch?v=ZZ7oht6TD9c
    - philipallstar 2mo ago
      The sinister nature of that integration doesn't make this one sinister at all.
      - AnthonyMouse 2mo ago
        Do you expect Hugging Face to remain a neutral platform with no preference for any given GPU vendor's technology?
    - huurtehoog 2mo ago
      That's not vertical integration though. Vertical integration would be the corn chip maker buying corn farms for example. I don't think there's a name for "buying makers of complementary goods".
      - micromacrofoot 2mo ago
        vertical integration
        huurtehoog 2mo ago
        "In microeconomics, management and international political economy, vertical integration, also referred to as vertical consolidation, is an arrangement in which the supply chain of a company is integrated and owned by that company." [emphasis mine] https://en.wikipedia.org/wiki/Vertical_integration https://en.wikipedia.org/wiki/Vertical_integration
        micromacrofoot 2mo ago
        horizontal integration
        huurtehoog 2mo ago
        "Horizontal integration is the process of a company increasing production of goods or services at the same level of the value chain, in the same industry. A company may do this via internal expansion or through mergers and acquisition" [emphasis mine] https://en.wikipedia.org/wiki/Horizontal_integration https://en.wikipedia.org/wiki/Horizontal_integration Anything else you care to throw at the wall to see if it sticks, just go ahead and assume I disagree with you given how our interaction has gone so far.
bingemaker 2mo ago
> Nvidia has a strong interest in open source Linus would heavily disagree!
- therealmarv 2mo ago
  He said 2 years ago they are okay nowadays (they were not in the past). source: https://diginomica.com/kubecon-china-33-and-third-linux-long-player-so-why-does-linus-torvalds-hate-ai https://diginomica.com/kubecon-china-33-and-third-linux-long...
  - croes 2mo ago
    But he also says why, so he knows as soon as Linux isn't useful for Nvidias revenue anymore they will be gone
    - xienze 2mo ago
      All the big boy AI stuff runs on servers, which certainly aren't going to be running on Windows or Mac anytime soon. Linux _is_ what all this NVIDIA hardware is going to run on until... what exactly? FreeBSD overtakes Linux in the server space?
      - croes 2mo ago
        Who knows what i sthe next big thing after AI. Maybe something that doesn't need Linux? And I'm not so sure that Mac is out of the equation
        newswasboring 2mo ago
        Can we at least let the early days of AI play out before we try to guess then next big thing? ChatGPT isn't even half a decade old.
        croes 2mo ago
pelagicAustral 2mo ago
NVidia has a strong interest in profit. The only reason we are allowed to live in chip shangri-la is because they can make a profit out of it, nothing else.
- andybak 2mo ago
  Nothing you wrote contradicts the post you are replying to.
  - pelagicAustral 2mo ago
    I agree with you
globalise83 2mo ago
Yes, this strategy is known as "commoditize your complement"
lofaszvanitt 2mo ago
Open source models are only there so the closed ones wouldn't be attacked from every angle.
j-bos 2mo ago
But Nvidia isn't the only gpu maker..,
croes 2mo ago
As long as the models run best on Nvidia hardware. What will happen if models are faster on Apple or AMD hardware?
- Danox 2mo ago
  Prediction Apple will find the exit…
- j45 2mo ago
  Nvidia being making it clear that other hardware models will permanently not become second class citizens in any way would be a great start.
  - croes 2mo ago
    Talk is cheap, so no need to claim that in the first place. Of course they could make it legally binding but they won't do it. They will be as free as the Washington Post after the aquisition by Bezos. Wait and see is the only thing we can do.
    - Danox 2mo ago
      You haven’t been paying attention to Washington Post isn’t free to do anything they want under Jeff Bezos…
  - Const-me 2mo ago
    For the last few months, I’ve been working on a vendor-agnostic inference library. Have 3 backends so far: legacy D3D11 for compatibility, D3D12, and Vulkan 1.3. I have reasons to believe nVidia deliberately crippling Vulkan API for their consumer GPUs. Couple examples to be specific. nVidia driver sets quite low number for VkPhysicalDeviceLimits::maxTexelBufferElements. I don’t think that’s a hardware limit because I have D3D12 backend doing the same thing on the same hardware. Vulkan performance is not great on nVidia. On all AMD cards I am testing, Vulkan 1.3 is the fastest backend. On nVidia however, D3D12 is faster despite tensor cores (WMMA / wave matrix multiply accumulate / cooperative matrices) are only available through Vulkan, D3D12 is using shader cores exclusively.
ggvnfnjff 2mo ago
Open AI's jalapeño would be an example of this exactly...
taneq 2mo ago
If you wanna sell pickaxes to miners, you want a lot of independent miners doing well.
PunchyHamster 2mo ago
MS also have interest. The problem is that microsoft is fucking incompetent at running it. Not sure if NVIDIA is any better on that front

fjcururuvy7 2mo ago

HN failing to understand embrace extend extinguish yet again I see.

hanselot 2mo ago
EEE is impossible already. They can't close pandoras box now. Why is it alarming for a shovel salesman to also sell plants?
- AnthonyMouse 2mo ago
  Context: Liz Lemon has been buying two products together because one of them creates a need for the other. Jack Donaghy: The only thing I will be discussing with the House Subcommittee on Baseball, Quiz Shows, Terrorism, and Media is vertical integration. Liz Lemon: What's vertical integration? Jack Donaghy: Imagine that your favorite corn chip manufacturer also owned the number one diarrhea medication. Liz Lemon: That'd be great, 'cause then they could put a little sample of the medicine in each bag. Jack Donaghy: Keep thinking. Liz Lemon: [beat] Except then they might be tempted to make the corn chips give you... Jack Donaghy: Vertical integration. https://www.youtube.com/watch?v=ZZ7oht6TD9c https://www.youtube.com/watch?v=ZZ7oht6TD9c
  - philipallstar 2mo ago
    The sinister nature of that integration doesn't make this one sinister at all.
    - AnthonyMouse 2mo ago
      Do you expect Hugging Face to remain a neutral platform with no preference for any given GPU vendor's technology?
  - huurtehoog 2mo ago
    That's not vertical integration though. Vertical integration would be the corn chip maker buying corn farms for example. I don't think there's a name for "buying makers of complementary goods".
    - micromacrofoot 2mo ago
      vertical integration
      - huurtehoog 2mo ago
        "In microeconomics, management and international political economy, vertical integration, also referred to as vertical consolidation, is an arrangement in which the supply chain of a company is integrated and owned by that company." [emphasis mine] https://en.wikipedia.org/wiki/Vertical_integration https://en.wikipedia.org/wiki/Vertical_integration
        micromacrofoot 2mo ago
        horizontal integration
        huurtehoog 2mo ago
        "Horizontal integration is the process of a company increasing production of goods or services at the same level of the value chain, in the same industry. A company may do this via internal expansion or through mergers and acquisition" [emphasis mine] https://en.wikipedia.org/wiki/Horizontal_integration https://en.wikipedia.org/wiki/Horizontal_integration Anything else you care to throw at the wall to see if it sticks, just go ahead and assume I disagree with you given how our interaction has gone so far.

bingemaker 2mo ago

> Nvidia has a strong interest in open source Linus would heavily disagree!

therealmarv 2mo ago
He said 2 years ago they are okay nowadays (they were not in the past). source: https://diginomica.com/kubecon-china-33-and-third-linux-long-player-so-why-does-linus-torvalds-hate-ai https://diginomica.com/kubecon-china-33-and-third-linux-long...
- croes 2mo ago
  But he also says why, so he knows as soon as Linux isn't useful for Nvidias revenue anymore they will be gone
  - xienze 2mo ago
    All the big boy AI stuff runs on servers, which certainly aren't going to be running on Windows or Mac anytime soon. Linux _is_ what all this NVIDIA hardware is going to run on until... what exactly? FreeBSD overtakes Linux in the server space?
    - croes 2mo ago
      Who knows what i sthe next big thing after AI. Maybe something that doesn't need Linux? And I'm not so sure that Mac is out of the equation
      - newswasboring 2mo ago
        Can we at least let the early days of AI play out before we try to guess then next big thing? ChatGPT isn't even half a decade old.
        croes 2mo ago

pelagicAustral 2mo ago

NVidia has a strong interest in profit. The only reason we are allowed to live in chip shangri-la is because they can make a profit out of it, nothing else.

andybak 2mo ago
Nothing you wrote contradicts the post you are replying to.
- pelagicAustral 2mo ago
  I agree with you

globalise83 2mo ago

Yes, this strategy is known as "commoditize your complement"

lofaszvanitt 2mo ago

Open source models are only there so the closed ones wouldn't be attacked from every angle.

j-bos 2mo ago

But Nvidia isn't the only gpu maker..,

croes 2mo ago

As long as the models run best on Nvidia hardware. What will happen if models are faster on Apple or AMD hardware?

Danox 2mo ago
Prediction Apple will find the exit…
j45 2mo ago
Nvidia being making it clear that other hardware models will permanently not become second class citizens in any way would be a great start.
- croes 2mo ago
  Talk is cheap, so no need to claim that in the first place. Of course they could make it legally binding but they won't do it. They will be as free as the Washington Post after the aquisition by Bezos. Wait and see is the only thing we can do.
  - Danox 2mo ago
    You haven’t been paying attention to Washington Post isn’t free to do anything they want under Jeff Bezos…
- Const-me 2mo ago
  For the last few months, I’ve been working on a vendor-agnostic inference library. Have 3 backends so far: legacy D3D11 for compatibility, D3D12, and Vulkan 1.3. I have reasons to believe nVidia deliberately crippling Vulkan API for their consumer GPUs. Couple examples to be specific. nVidia driver sets quite low number for VkPhysicalDeviceLimits::maxTexelBufferElements. I don’t think that’s a hardware limit because I have D3D12 backend doing the same thing on the same hardware. Vulkan performance is not great on nVidia. On all AMD cards I am testing, Vulkan 1.3 is the fastest backend. On nVidia however, D3D12 is faster despite tensor cores (WMMA / wave matrix multiply accumulate / cooperative matrices) are only available through Vulkan, D3D12 is using shader cores exclusively.

ggvnfnjff 2mo ago

Open AI's jalapeño would be an example of this exactly...

taneq 2mo ago

If you wanna sell pickaxes to miners, you want a lot of independent miners doing well.

PunchyHamster 2mo ago

MS also have interest. The problem is that microsoft is fucking incompetent at running it. Not sure if NVIDIA is any better on that front