Nvidia bans using translation layers for CUDA software
Nvidia bans using translation layers for CUDA software
Translators in the crosshairs.
![Nvidia bans using translation layers for CUDA software — previously the prohibition was only listed in the online EULA, now included in installed files [Updated]](https://lemmy.world/pictrs/image/cbb01be6-711d-464f-a638-c9dfe142bb5b.jpeg?format=webp&thumbnail=256)
Nowadays you cant do anything with the software or hardware you put and have on your pc.
If nvidia is going to go on a power trip, then please make that nvidia drivers is only allowed to get installed by nvidia servicemen before that the servicemen teaches the user about their 30 thousand page eula what and what they can do with THEIR bought hardware.
Maybe we should rent our video cards for $25 per month. You get 2,000,000 frames rendered per month and anything beyond that puts you in a pro gamer tier for more money.
It'll be a collab with HP!
heh, if ye had yer screen on 24/7 that would be merely 0.83 frames per second
The human eye can't see more than 0.5 frames per second anyways (/s)
and what they
canare allowed by Nvidia to do with THEIR bought hardware.I read the article, and a few points stuck out to me:
- This has been a restriction since 2021; now it’s documented in the files and not just the online EULA (ie consistent)
- This is a protection to disallow other companies like Intel and AMD from profiting off of Nvidia’s work
- Nothing is stopping anybody from porting the software to other hardware, eg
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.
I’m all for piracy and personal freedoms, but it doesn’t seem to be what this is about. It’s about combating other companies profiting off Nvidia’s work. Companies should be able to fight back against other companies (or countries).
I mean it’s not like Nvidia is unreasonably suing open-source projects into oblivion or anything, or subpoenaing websites for user data; at least, not yet.
There's a good argument that Nvidia only had the money to do the work because of anticompetitive practices, and so shouldn't be allowed to benefit from it unless everyone's allowed to benefit from it, otherwise it's just cementing their dominant position further.
"not yet" is carrying a mighty load of hope there. Monopolists are going to monopolize.
Thanks for this rational breakdown of what's actually happening. Pretty misleading headline tbh.
None of their business if we use a translation layer.
Agreed. I hope lawmakers step in and make EULAs like this unenforceable.
EULAs are already by definition unenforceable in the EU
This feels illegal. Like it's probably not, but it should be.
It probably is. In the EU APIs aren't copyrightable in the first place, doubly so if it's necessary for interoperability, in the US there's Google vs. Oracle which declared Google's use of Java APIs in Android fair use.
That's the neat thing about being in the American oligarchs class. If it's illegal just make it legal.
Feels like a fantastic base for an anti-trust case at the least.
Definitely not anticompetitive in the slightest. No need to look here, FTC.
An anti-trust lawsuit is overdue
Nvidia is dominating the AI chip market. If our laws were properly enforced, Nvidia should’ve been too afraid to abuse their market position like this.
I guess this is Nvidia's reaction to projects like ZLUDA.
And that's a textbook case why monopolies are bad for pretty much everyone except the shareholders of that monopolistic company.
I am extremely tempted to @ some guy who was shilling for nvidia and saying they weren't a monopoly.
Part of it depends on how you define things. They're not a monopoly in terms of having eliminated all their competitors, but they're a defacto monopoly in terms of being able to do the things a monopoly can. As an example, they can dictate pricing for the whole market as their margins are better than AMD's, so if AMD undercut them, they can retaliate by dropping their prices to the point AMD would have to sell at cost, so AMD can only sell things in the narrow price window where Nvidia doesn't feel threatened. On the other hand, AMD does exist and does sell things.
In gaming subs you'll meet them plenty
Is something like this actually enforceable? That's like Microsoft saying you can't use Wine on Linux.
Wine is done on clean room reverse engineering, it doesn't use any propetriary code as reference. If they had done so, Microsoft would have grounds to sue them.
This can't enforce anything on CUDA versions below 11.6; but any functionality introduced to CUDA after 11.6 needs to be clean room reverse engineered, so this will make ZLUDA development on those versions more difficult.
Yeah, Wine is very strict about this; IIRC if you've ever even looked at the leaked Windows XP source code, you're not allowed to work on Wine.
It might not be. But the mere potential of having to litigate for years will have a chilling effect.
Interoperability is illegal now?
Take a page from the AI companies' book - just claim AI "learned" from the CUDA SDK and call it fair use.
Yes, the clause might be unenforceable on fair use grounds. So, if you feel like going through a couple years of risky litigation...
Funny how people aren't cheering on NVIDIA.
Is this in response to AMD's cuda adaptor
I tried to read the article but i am too stupid. I think nvidia has a proprietary hardware/software combo that is very fast, but because they "own it" they want money; instead other companies are using this without paying... Am i close?
You can use graphics cards for more than just graphics, eg for AI. Nvidia is a leader in facilitating that.
They offer a software toolkit for developing programs (an SDK) that use their GPUs to best effect. People have begun making "translation layers" that allow such CUDA programs to run on non-nvidia hardware. (I have no idea how any of this works.) The license of that SDK now forbids reverse engineering its output to create these compatibility tools.
Unless I am very mistaken, Nvidia can't ban the use of "translation layers" or stop people making them, as such. This clause creates a barrier to creating them, though.
Some programs will probably remain CUDA specific, because of that clause. That means that Nvidia is a gatekeeper for these programs and can charge extra for access.
Thanks
It's not about it being fast, it's about it only being available for NVidia GPUs. As long as software for things like machine learning uses CUDA, you need to buy an NVidia GPU to use it. A translation layer would let you use the same software on other companies' GPUs, which means people aren't forced to buy NVidia's GPUs anymore.
How does this make sense? If you've got an NVIDIA card, you don't need an emulation level. And if you have a different hardware that needs an emulation layer, you don't have to agree to those NVIDIA terms, because you are not using their products.
Cuda is the main reason Nvidia has their monopoly. Especially their artifiical limitations on VRAM for more expensive cards would make AMD a lot more interesting if AMD actually had good support.
Is this not similar to the Android Java interface?