3 ms·

If AMD can release solid hardware, all they would need to do is add support for their hardware to popular open-source projects like TensorFlow. I just hope the

by mabbo 10y ago

If AMD can release solid hardware, all they would need to do is add support for their hardware to popular open-source projects like TensorFlow. I just hope they do the second part correctly.

AlphaSite 10y ago
It looks like they're doing it, or already have done so. https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/45 https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/iss...
- RockyMcNuts 10y ago
  TensorFlow is evolving rapidly and it's a challenge to keep up with GPU driver updates, CUDA updates, TensorFlow updates, and not fall into version hell. For instance I took the Udemy course a few months back, given by a Google scientist, and in at least one instance the code was pinned to CPU because it didn't run on GPU, presumably due to a TensorFlow issue. If they can get TensorFlow running well on AMD and release images that make it painless, it will be a huge win, but I'm a little skeptical. Big difference between getting a benchmark win and a stable platform.
devy 10y ago
FYI, AMD's people are helping TF to add OpenCL support. https://github.com/tensorflow/tensorflow/issues/22 https://github.com/tensorflow/tensorflow/issues/22
- dharma1 10y ago
  Contributions from AMD are sorely lacking. It's mostly people from https://www.codeplay.com/ https://www.codeplay.com/ and https://github.com/hughperkins/ https://github.com/hughperkins/
  - devy 10y ago
    I noticed Hugh Perkins has a AMD badge under "Organizations". Is he not working for AMD? Also one of the earlier participants of the TF OpenCL support conversations https://github.com/gujunli https://github.com/gujunli was working for AMD at the time[1]. https://electrek.co/2016/02/26/tesla-machine-learning-expert-autopilot/ https://electrek.co/2016/02/26/tesla-machine-learning-expert...
    - dharma1 10y ago
      No, Hugh works at asapp.com. AMD could really invest more in optimising deep learning libraries for their hardware if they want to be relevant
      - devy 10y ago
        Interesting! He's definitely got some interesting background there. Normally someone wouldn't be added into a Github organization without an employment relationship there. > AMD could really invest more in optimising deep learning libraries for their hardware if they want to be relevant I am optimistic about AMD's involvement in AI. Just do a search of recent news on here, tons of exciting devleopment from AMD, especially Radeon Instinct. And here is an interesting comment about the comparison between Nvidia and AMD[1]. [1] https://news.ycombinator.com/item?id=13161050 https://news.ycombinator.com/item?id=13161050

by mabbo 10y ago

If AMD can release solid hardware, all they would need to do is add support for their hardware to popular open-source projects like TensorFlow. I just hope they do the second part correctly.

AlphaSite 10y ago

It looks like they're doing it, or already have done so. https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/45 https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/iss...

RockyMcNuts 10y ago
TensorFlow is evolving rapidly and it's a challenge to keep up with GPU driver updates, CUDA updates, TensorFlow updates, and not fall into version hell. For instance I took the Udemy course a few months back, given by a Google scientist, and in at least one instance the code was pinned to CPU because it didn't run on GPU, presumably due to a TensorFlow issue. If they can get TensorFlow running well on AMD and release images that make it painless, it will be a huge win, but I'm a little skeptical. Big difference between getting a benchmark win and a stable platform.

devy 10y ago

FYI, AMD's people are helping TF to add OpenCL support. https://github.com/tensorflow/tensorflow/issues/22 https://github.com/tensorflow/tensorflow/issues/22

dharma1 10y ago
Contributions from AMD are sorely lacking. It's mostly people from https://www.codeplay.com/ https://www.codeplay.com/ and https://github.com/hughperkins/ https://github.com/hughperkins/
- devy 10y ago
  I noticed Hugh Perkins has a AMD badge under "Organizations". Is he not working for AMD? Also one of the earlier participants of the TF OpenCL support conversations https://github.com/gujunli https://github.com/gujunli was working for AMD at the time[1]. https://electrek.co/2016/02/26/tesla-machine-learning-expert-autopilot/ https://electrek.co/2016/02/26/tesla-machine-learning-expert...
  - dharma1 10y ago
    No, Hugh works at asapp.com. AMD could really invest more in optimising deep learning libraries for their hardware if they want to be relevant
    - devy 10y ago
      Interesting! He's definitely got some interesting background there. Normally someone wouldn't be added into a Github organization without an employment relationship there. > AMD could really invest more in optimising deep learning libraries for their hardware if they want to be relevant I am optimistic about AMD's involvement in AI. Just do a search of recent news on here, tons of exciting devleopment from AMD, especially Radeon Instinct. And here is an interesting comment about the comparison between Nvidia and AMD[1]. [1] https://news.ycombinator.com/item?id=13161050 https://news.ycombinator.com/item?id=13161050