Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Nvidia releases open-source GPU kernel modules

developer.nvidia.com

Post reply on HN
Print view View original

391–400 of 415 posts

Re: Nvidia releases open-source GPU kernel modules

#391

Post by worthless-trash »

prox wrote: post #346
loganberries wrote: post #331

Earlier quoted context omitted.

Currently, you rely on Nvidia to keep the drivers up to date and working with the kernel you use in your OS. If the drivers aren't upstream, a future kernel can break them. So if you update your OS, it might no longer support your GPU without installing an alternative. Nvidia are responsible for getting this fixed in a timely manner. If the drivers are upstream, then people aren't allowed to break them when they make…

Thanks, so it means a better out of the box stability for end users.

And when you update your kernel, you wont need to recompile the kernel modules every time.

Re: Nvidia releases open-source GPU kernel modules

#392

Post by IntelMiner »

cvuls wrote: post #288
hyperionplays wrote: post #278

Earlier quoted context omitted.

this was the case with RTX voice, a simple regedit made it work on non-RTX cards fine.

From my understanding, it used CPU acceleration instead

It would use RT cores by default, but if they weren't available or in use (EG: running a game with Ray Tracing) it could fall back to the older CUDA

That's also why they split it out into two products now

"Nvidia Broadcast" is their "you bought an RTX GPU so here's a shiny toy"

And "RTX Voice" for people with GPU's dating back to the Geforce 400 series that only does it via CUDA

Re: Nvidia releases open-source GPU kernel modules

#393

Post by native_samples »

KingMachiavelli wrote: post #337
native_samples wrote: post #326

Earlier quoted context omitted.

Drivers are a large part of the development costs of a GPU, just like how an operating system is a large part of the development cost of a general computer. If nVidia open sourced all their drivers tomorrow the risk of some cheap Chinese shop making clones of their hardware and re-targeting the nVidia drivers to get lots of the features would be very high. It'd significantly reduce the value of what they'd built (to…

AMD and Intel both have open source video drivers and I really doubt anyone has cloned them. The non-free features for the Nvidia cards and many other chips are achieved by the card itself running closed source binaries. It is actually probably a better way to protect IP anyway since no one can decompile the encrypted binary (well until recently...). Open source and proprietary software exist as a duality. It's not a…

AMD and Intel aren't (or weren't at least) shipping cutting edge GPU tech so there's less need to look to them for ideas and shortcuts anyway.

Well I used to work at Google and the open source parts are a tiny, tiny fraction of their actual codebase. Windows isn't really open source at all, although in recent years a few utilities have been opened up - note, only after Windows stopped being so important to Microsoft.

And as for macOS. Well. You can download and read some code, sometimes. Good luck trying to actually build it or do anything useful with it at all. You'll find that it's (a) completely undocumented and (b) all depends on internal stuff you don't have access to. An exception is WebKit.

Re: Nvidia releases open-source GPU kernel modules

#394

Post by native_samples »

anthk wrote: post #381
native_samples wrote: post #326

Earlier quoted context omitted.

Drivers are a large part of the development costs of a GPU, just like how an operating system is a large part of the development cost of a general computer. If nVidia open sourced all their drivers tomorrow the risk of some cheap Chinese shop making clones of their hardware and re-targeting the nVidia drivers to get lots of the features would be very high. It'd significantly reduce the value of what they'd built (to…

You mean, like Intel GPU drivers with just the libre free-as-in-freedom kernel and MESA? Blender? Krita? Cinelerra-CV? Darktable? ImageMagick? FFMPEG? KVM/Qemu? Most programming languages and frameworks? Clang/LLVM? WTF are you talking about? If any, science today it's made thanks to FLOSS software, propietary software it's the exception . And the trend it looking worse for propietary environments. Money will come fr…

Krita? You're just naming random open source Linux programs that hardly anyone uses.

Yes, we can all make long lists of open source projects. That's not my point. The point is that the average person, on an average day, is using lots of proprietary software (or proprietary forks of open source software).

I've been hearing about how the future of software is charging for support tickets and hand-waved 'integration' for 30 years. The biggest, richest and most powerful tech firms today all ignored that advice. There's only one company that did well out of that approach and they're now called IBM.

Re: Nvidia releases open-source GPU kernel modules

#395

Post by native_samples »

account42 wrote: post #350
native_samples wrote: post #324

Earlier quoted context omitted.

Yes. Drivers represent enormous amounts of work and large parts of the overall development/IP effort in a GPU. Asking why nVidia don't open source it is like asking why they don't just open source the entire GPU. Answer: because then they'd have a much smaller and less valuable business, if they even had a business at all.

> because then they'd have a much smaller and less valuable business, if they even had a business at all. This does not follow. While GPU drivers are a huge development effort, they are also very specific to the hardware. This is even more true with modern APIs that are closer to how the hardware works. AMD and Intel both have open source GPU drivers and they are still in business. Are you claiming that their busines…

Apparently nVidia believe that is the case, and they are the market leaders, so - yes?

Re: Nvidia releases open-source GPU kernel modules

#396

Post by aspaceman »

sylware wrote: post #18

The kernel modules are the first stage I guess, since a massive amount of hardware programing knowledge is in user space like with AMD/intel GPUs. I wonder how much LAPSUS$ hack has to do with it. I wonder if nvidia hardware programing interface is a mess like AMD one, just curious.

Do you have a reference for the AMD interface? I know it exists but don't know where to find it.

Re: Nvidia releases open-source GPU kernel modules

#399

Post by azeirah »

prash_ant wrote: post #271
greggsy wrote: post #266

Earlier quoted context omitted.

Has any rationale been provided around the blob being closed? Is it licensed, or simply too too complicated to pick apart? Graphics is a complicated field, but the techniques could still be patented - surely their market status isn’t dependent on trade secrets?

There have been reports earlier [1] that processing power of some GPUs is suppressed by software rather than the hardware capability itself. It is frequently easier to mass produce similar chips than to have different chips for different priced devices. I had come across comments in other online forums where the users alleged that some software flags restrict the capability. I hope someone else will link those webpag…

I suppose this is also used for yield quality?

For instance, AMD is known to sell the same actually-8-core chip as a 6-core if the transistor yield was poorer.

I believe this is a standard in the industry, right? You can theoretically unlock these cores yourself, but there's a decent chance it will break or cause other significant problems.

Re: Nvidia releases open-source GPU kernel modules

#400

Post by mjan22640 »

hoistbypetard wrote: post #354
adrian_b wrote: post #347

Earlier quoted context omitted.

All closed-source kernel modules and a part of the firmware blobs (those that are for devices which have communication interfaces, especially wireless interfaces, and which may access directly the memory space without being blocked by a MMU under the control of the OS kernel) are major security risks, because they might be controlled by an attacker, with no means of detection by the user. With a security-strengthened…

> With a security-strengthened operating system, or with a VM/container/sandbox, it is possible to ensure that any opaque user-space program is no security risk. In theory this may be true. But the particular opaque user-space program is talking to a driver authored by the same team. That driver is talking to hardware that has access to all of the memory in your computer. The odds are high that a userspace program ta…

> has access to all of the memory in your computer

You can use IOMMU to restrict the access to the appropriate memory sections.

Post reply on HN