5 ms·

> The problem is if I spend an evening trying to do anything with OpenCL or ROCm the kernel hard-locks and I go to bed early. In a race against NVIDIA where ev

> The problem is if I spend an evening trying to do anything with OpenCL or ROCm the kernel hard-locks and I go to bed early. In a race against NVIDIA where every comment on here is about just how much better CUDA is than any alternative, why doesn't literally what you just said "also" have the attention of Lisa Su?

georgehotz 3y ago
I have had multiple replies from Lisa Su on both Twitter and by e-mail. It doesn't help. AMD is structurally incapable of fixing these issues. They don't even have a 7900XTX in CI. Crashing the firmware is so trivial there's no way anyone fuzzed anything. They debug by the application, adding mitigations at many layers of the stack to make it work. While this strategy is fine for the 20 mainstream games that come out per year, unless you root cause issues and have a good CI, you'll never build a stable GPU for general compute.
- MuffinFlavored 3y ago
  Genuine question for you then, why are you investing your time reversing their firmware if it sounds like you think they're failing/will fail? Why not just "adopt" the golden child/golden standard NVIDIA now for your usecase, reverse that (if needed) and move on? Genuinely curious what you gain from trying to make AMD work/investing in AMD on your end here. Is it that you see a value proposition through the "muck" of crashy-kernels that I'm missing?
  - georgehotz 3y ago
    This is my weekend hobby, I'm learning about GPUs. Besides, I'd like to see competition to NVIDIA. From a business side, tiny corp is selling boxes with both AMD and NVIDIA, we'll let the user choose how much they value money vs pain.
    - rudedogg 3y ago
      What are your thoughts on using Intel GPUs for deep learning?
    - doctorpangloss 3y ago
      Did you ever consider using NVIDIA? I thought the machine you were planning to sell originally had RTX 6000s. Also, isn’t the real problem with the 7900 XTX is the lack of interconnect? I really appreciate the long list of stuff you are agitating AMD about. Infinity Fabric and blower fans would be great too. Another POV is none of that is ever going to happen, and everyone will have to wait for a Chinese GPU manufacturer to create real competition. I can’t recall a single time in my life a non-Chinese company - electronics or otherwise - delivered the “same” features for a lower price.
    - chem83 3y ago
      Thoughts on IREE[0]'s ability to close this gap? [0] iree.dev
  - braiamp 3y ago
    > Why not just "adopt" the golden child/golden standard NVIDIA now for your usecase, reverse that (if needed) and move on? Oh oh, I will answer you with the business side of the thing: because Nvidia sells exactly the same thing as Tinycorp. Also, Nvidia has shown that they will crush you if they get a whiff that you are doing something they don't like, for example, speaking with a competitor. With a business partner that gets you by the balls like that, it is not sound business strategy not to depend on them.
- jimbob45 3y ago
  They don't even have a 7900XTX in CI. You read stories in the glory days of Microsoft where they would bend over backwards to ensure backwards compatibility [0]. Going from that to what you say is depressing. It’s actually hard for me to accept your statement because such a failure would not only require QA/the PM to be totally ignorant of current testing best practices, but also require HR to have catastrophically misidentified both those people and the people who supervise them. Keep up the great work. [0]https://devblogs.microsoft.com/oldnewthing/20191119-00/?p=103115 https://devblogs.microsoft.com/oldnewthing/20191119-00/?p=10...
  - llm_trw 3y ago
    The shocking thing about AMD in the last 5 years isn't that they got Ryzen to work, it's that they didn't screw it up within two releases. There are no other technology companies in the world who have a will to fail so ingrained in their corporate DNA as does AMD.
- lofaszvanitt 3y ago
  Maybe they are not allowed to compete...
- makomk 3y ago
  Yikes. For context, the 7900XTX and 7900XT are the only consumer GPUs that AMD officially support running AI workloads on, and the XTX is the one they recommend out of the two. Other cards in the same generation and any from previous generations are officially not supported or tested and may not work. So it's not like they have a particularly large amount of hardware that needs to be included in their testing environment.

georgehotz 3y ago

I have had multiple replies from Lisa Su on both Twitter and by e-mail. It doesn't help. AMD is structurally incapable of fixing these issues. They don't even have a 7900XTX in CI. Crashing the firmware is so trivial there's no way anyone fuzzed anything. They debug by the application, adding mitigations at many layers of the stack to make it work. While this strategy is fine for the 20 mainstream games that come out per year, unless you root cause issues and have a good CI, you'll never build a stable GPU for general compute.

MuffinFlavored 3y ago
Genuine question for you then, why are you investing your time reversing their firmware if it sounds like you think they're failing/will fail? Why not just "adopt" the golden child/golden standard NVIDIA now for your usecase, reverse that (if needed) and move on? Genuinely curious what you gain from trying to make AMD work/investing in AMD on your end here. Is it that you see a value proposition through the "muck" of crashy-kernels that I'm missing?
- georgehotz 3y ago
  This is my weekend hobby, I'm learning about GPUs. Besides, I'd like to see competition to NVIDIA. From a business side, tiny corp is selling boxes with both AMD and NVIDIA, we'll let the user choose how much they value money vs pain.
  - rudedogg 3y ago
    What are your thoughts on using Intel GPUs for deep learning?
  - doctorpangloss 3y ago
    Did you ever consider using NVIDIA? I thought the machine you were planning to sell originally had RTX 6000s. Also, isn’t the real problem with the 7900 XTX is the lack of interconnect? I really appreciate the long list of stuff you are agitating AMD about. Infinity Fabric and blower fans would be great too. Another POV is none of that is ever going to happen, and everyone will have to wait for a Chinese GPU manufacturer to create real competition. I can’t recall a single time in my life a non-Chinese company - electronics or otherwise - delivered the “same” features for a lower price.
  - chem83 3y ago
    Thoughts on IREE[0]'s ability to close this gap? [0] iree.dev
- braiamp 3y ago
  > Why not just "adopt" the golden child/golden standard NVIDIA now for your usecase, reverse that (if needed) and move on? Oh oh, I will answer you with the business side of the thing: because Nvidia sells exactly the same thing as Tinycorp. Also, Nvidia has shown that they will crush you if they get a whiff that you are doing something they don't like, for example, speaking with a competitor. With a business partner that gets you by the balls like that, it is not sound business strategy not to depend on them.
jimbob45 3y ago
They don't even have a 7900XTX in CI. You read stories in the glory days of Microsoft where they would bend over backwards to ensure backwards compatibility [0]. Going from that to what you say is depressing. It’s actually hard for me to accept your statement because such a failure would not only require QA/the PM to be totally ignorant of current testing best practices, but also require HR to have catastrophically misidentified both those people and the people who supervise them. Keep up the great work. [0]https://devblogs.microsoft.com/oldnewthing/20191119-00/?p=103115 https://devblogs.microsoft.com/oldnewthing/20191119-00/?p=10...
- llm_trw 3y ago
  The shocking thing about AMD in the last 5 years isn't that they got Ryzen to work, it's that they didn't screw it up within two releases. There are no other technology companies in the world who have a will to fail so ingrained in their corporate DNA as does AMD.
lofaszvanitt 3y ago
Maybe they are not allowed to compete...
makomk 3y ago
Yikes. For context, the 7900XTX and 7900XT are the only consumer GPUs that AMD officially support running AI workloads on, and the XTX is the one they recommend out of the two. Other cards in the same generation and any from previous generations are officially not supported or tested and may not work. So it's not like they have a particularly large amount of hardware that needs to be included in their testing environment.