Nvidia bans using translation layers for CUDA software — previously the prohibition was only listed in the online EULA, now included in installed files [Updated]

(Image credit: AMD)

[Edit 3/4/24 11:30am PT: Clarified article to reflect that this clause is available on the online listing of Nvidia's EULA, but has not been in the EULA text file included in the downloaded software. The warning text was added to 11.6 and newer versions of the installed CUDA documentation.]

Nvidia has banned running CUDA-based software on other hardware platforms using translation layers in its licensing terms listed online since 2021, but the warning previously wasn't included in the documentation placed on a host system during the installation process. This language has been added to the EULA that's included when installing CUDA 11.6 and newer versions.

The restriction appears to be designed to prevent initiatives like ZLUDA, which both Intel and AMD have recently participated, and, perhaps more critically, some Chinese GPU makers from utilizing CUDA code with translation layers. We've pinged Nvidia for comment and will update you with additional details or clarifications when we get a response.

Longhorn, a software engineer, noticed the terms. "You may not reverse engineer, decompile or disassemble any portion of the output generated using SDK elements for the purpose of translating such output artifacts to target a non-NVIDIA platform.," a clause in the installed EULA text file reads.

The clause was absent in the EULA documentation that's installed with the CUDA 11.4 and 11.5 release, and presumably with all versions before that. However, it is present in the installed documentation with version 11.6 and newer.

Being a leader has a good side and a bad side. On the one hand, everyone depends on you; on the other hand, everyone wants to stand on your shoulders. The latter is apparently what has happened with CUDA. Because the combination of CUDA and Nvidia hardware has proven to be incredibly efficient, tons of programs rely on it. However, as more competitive hardware enters the market, more users are inclined to run their CUDA programs on competing platforms. There are two ways to do it: recompile the code (available to developers of the respective programs) or use a translation layer.

For obvious reasons, using a translation layer like ZLUDA is the easiest way to run a CUDA program on non-Nvidia hardware. All one has to do is take already-compiled binaries and run them using ZLUDA or other translation layers. ZLUDA appears to be floundering now, with both AMD and Intel having passed on the opportunity to develop it further, but that doesn't mean translation isn't viable.

Several Chinese GPU makers, including one funded by the Chinese government, claim to run CUDA code. Denglin Technology designs processors featuring a "computing architecture compatible with programming models like CUDA/OpenCL." Given that reverse engineering of an Nvidia GPU is hard (unless one already somehow has all the low-level details about Nvidia GPU architectures), we are probably dealing with some sort of translation layer here, too.

One of the largest Chinese GPU makers, Moore Threads, also has a MUSIFY translation tool designed to allow CUDA code to work with its GPUs. However, whether or not MUSIFY falls under the classification of a complete translation layer remains to be seen (some of the aspects of MUSIFY could involve porting code). As such, it isn't entirely clear if the Nvidia ban on translation layers is a direct response to these initiatives or a pre-emptive strike against future developments.

For obvious reasons, using translation layers threatens Nvidia's hegemony in the accelerated computing space, particularly with AI applications. This is probably the impetus behind Nvidia's decision to ban running their CUDA applications on other hardware platforms using translation layers.

Recompiling existing CUDA programs remains perfectly legal. To simplify this, both AMD and Intel have tools to port CUDA programs to their ROCm (1) and OpenAPI platforms, respectively.

As AMD, Intel, Tenstorrent, and other companies develop better hardware, more software developers will be inclined to design for these platforms, and Nvidia's CUDA dominance could ease over time. Furthermore, programs specifically developed and compiled for particular processors will inevitably work better than software run via translation layers, which means better competitive positioning for AMD, Intel, Tenstorrent, and others against Nvidia — if they can get software developers on board. GPGPU remains an important and highly competitive arena, and we'll be keeping an eye on how the situation progresses in the future.

Anton Shilov
Freelance News Writer

Anton Shilov is a Freelance News Writer at Tom’s Hardware US. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.

  • Order 66
    This doesn't surprise me one bit. Tell me something surprising that Nvidia has done recently.
  • umeng2002_2
    The terms say "you." But it doesn't expressly prevent you from using translators made by someone else.
  • _sh4dow_
    Don't some countries, like Germany, have laws explicitly allowing reverse engineering for the purpose of achieving software compatibility in spite of any ToS restrictions?
  • ivan_vy
    "(unless one already has all the low-level details about Nvidia GPU architectures by stealing them)"
    not everything need to be stolen, poaching engineers is somewhat nefarious or gray area but kinda legal, ALL the companies do that, also people can choose who to work for, look at Jim Keller working with everyone and their neighbor.
    A little bit less of bias would be welcomed in the articles.
  • samopa
    Why one would use translator that possibly slower, and likely to be buggy, where one can recompile (not reprograms) their code to run natively, and possibly faster, on their target platform ?
  • AchakBrooks
    Nvidia is not the “root” in the hierarchy of “system”, yet they continually want to be this through their business practices. When they invest and become an “Apple type” then fine, be what they want.

    This type of limitation they want to set, kinda shows they are not confident in their hardware, and hints at their public sale pitch/reasoning to sell their hardware they way they do is just a grassy mountain built on a land fill.

    Honestly, maybe…. since they are building to be the “all in one/apple type” with their networking systems. Companies investing in such and seeing more publicly visible options, are more impowered to weigh those “options” based off the value of Nvidia’s “software” vs ”hardware” benefits over others that can harness the “software” side; would hurt Nvidia’s negotiation power no?

    PS: i have a passionate hate built up over decades towards everything Nvidia does so it’s hard to take a step back and give “the benefit of the doubt” or “this is the law so…”
  • Pierce2623
    The various governments of the world totally screwed up by letting the era of “you bought a piece of hardware and you bought a piece of software . Use them however you choose because they’re yours now” go away. Licensing on software and hardware ruined everything.
  • DiegoSynth
    samopa said:
    Why one would use translator that possibly slower, and likely to be buggy, where one can recompile (not reprograms) their code to run natively, and possibly faster, on their target platform ?
    Because developers will not do it. Specially huge corporations like 3D software makers.
    And because AMD / Intel users have the right to use their cards with these programs.
  • DiegoSynth
    Nvidia? The ones building for AI, which steals from creators? They are crying now?! Aw, maybe they need a bit of cream in their asses; must be irritated xD
  • NinoPino
    I agree with the conclusion of the article. Nvidia should open CUDA to other hardware and search for alliances, instead they closes it further. Imho is too late to use this strategy, especially after Intel consolidated Arc ecosystem and AMD proved MI300 capabilities and continued to improve ROCm.