RDNA 3 has a platform security processor?
Sounds like a stupid idea.
Documentation for the AMD 7900XTX
11–20 of 154 posts
Re: Documentation for the AMD 7900XTX
#12So for reference, this geohot is George Hotz who has a company Tiny Corp [0] in the space. Among a long list of things, he had a moment in the spotlight recently for a long rant [1] where he "gave up on AMD" because their drivers sucked more than he expected. The situation is reasonably complex - AMD have some programmers working on ROCm who seemed to be operating at standard fare when that is really not what AMD nee…
Anyone working on getting us out of the Nvidia monopoly on commerical AI should be praised. We need lots of competition in this space.
I'm hoping Google spins off its TPU division and sells chips to everyone that wants them.
I'm hoping AMD gets their act together and that they open up their drivers and stack.
I'm hoping a lot of other hardware companies and compute stacks start making inroads on this.
Nvidia is not just bilking the whole world on margin, but they're also a limiting reagent in the grand scheme of progress.
Re: Documentation for the AMD 7900XTX
#13So for reference, this geohot is George Hotz who has a company Tiny Corp [0] in the space. Among a long list of things, he had a moment in the spotlight recently for a long rant [1] where he "gave up on AMD" because their drivers sucked more than he expected. The situation is reasonably complex - AMD have some programmers working on ROCm who seemed to be operating at standard fare when that is really not what AMD nee…
I thought he sort of gave up and is going to build his stuff on Nvidia in the end? It's wild. I'd want AMD to present some sort of alternative if only because you can't get any high-mem chips these days but they seem to have some massive organizational uselessness on this front.
Re: Documentation for the AMD 7900XTX
#14I watched a bit of his attempts to reverse engineer the AMD GPU stack on YT, but Jesus is his keyboard loud.
did you learn anything?
Re: Documentation for the AMD 7900XTX
#15So for reference, this geohot is George Hotz who has a company Tiny Corp [0] in the space. Among a long list of things, he had a moment in the spotlight recently for a long rant [1] where he "gave up on AMD" because their drivers sucked more than he expected. The situation is reasonably complex - AMD have some programmers working on ROCm who seemed to be operating at standard fare when that is really not what AMD nee…
I thought he sort of gave up and is going to build his stuff on Nvidia in the end? It's wild. I'd want AMD to present some sort of alternative if only because you can't get any high-mem chips these days but they seem to have some massive organizational uselessness on this front.
They changed their mind a few days later and announced they're going to do NVIDIA and AMD versions of their box.
Re: Documentation for the AMD 7900XTX
#16We're talking about the company that couldn't fix a mouse cursor corruption bug affecting all of their GPU lines for what, 15, 20 years?
Re: Documentation for the AMD 7900XTX
#17RDNA 3 has a platform security processor? Sounds like a stupid idea.
Maybe that's where they run the HDCP/DRM stuff? That's the most obvious application of a secure enclave on a GPU that I can think of, and would also explain why they won't (or can't) open it up.
Re: Documentation for the AMD 7900XTX
#18This is quite crude on what it is. e.g. how does it differ from gpuopen[0] ISA documentation released by AMD themselves? 0. https://gpuopen.com/amd-isa-documentation/
The RDNA docs are just the tip of the iceberg.. in order to get to the compute engines (what you actually care about) you have to go through 3 layers of user-space cruft, a kernel driver, and then a firmware layer running half a dozen separate components of the GPU that "manage" the RDNA compute engines. Apparently, these components run on at least 3 different ISAs (some ARM, some F32, and some on RS64) where 2 of them aren't really documented at all. All of this is what he's trying to bypass and talk to the RDNA cores as directly as possible. Modern consumer GPUs are complex beasts.
Re: Documentation for the AMD 7900XTX
#19RDNA 3 has a platform security processor? Sounds like a stupid idea.
Maybe that's where they run the HDCP/DRM stuff? That's the most obvious application of a secure enclave on a GPU that I can think of, and would also explain why they won't (or can't) open it up.
Isn't HDCP/DRM implemented in ASIC hardware? There is nothing programmable there.