News for hackers. Stuff that matters.
Re:Tensorflow 2.0 AMD Support by zapnuk on Thursday November 28, 2019 @08:25AM (#21657107)
With nvidea not supporting MacOS beyond the current version, and Google pushing Swift for Tensorflow, I assume both Apple and Google have an increasing incentive to improve the adaptation of AMD GPUs for deep learing.
Re:Tensorflow 2.0 AMD Support by Vomzor on Thursday November 28, 2019 @10:52AM (#21658133)
Correct me if I'm wrong but I believe one of the advantages of macOS is that the GPU can use the system ram for deep learning. Which makes it possible to use cheaper GPUs. One of the reasons I'm looking forward to better AMD Tensorflow support.
Re:Tensorflow 2.0 AMD Support by p1esk on Friday November 29, 2019 @05:06AM (#21663382)
But you don't want to do that. The bandwidth between GPU and system RAM is 10x less than the bandwidth between GPU and GPU RAM, so the GPU would spend > 90% of time waiting for data. In this scenario, using CPU would most likely be faster.
Re:Tensorflow 2.0 AMD Support by nl on Thursday November 28, 2019 @06:17PM (#21660808)
Correct me if I'm wrong but I believe one of the advantages of macOS is that the GPU can use the system ram for deep learning. Which makes it possible to use cheaper GPUs.

I've never heard of this (and I do deep learning and use Macs). Do you have a reference?

Re:Tensorflow 2.0 AMD Support by girvo on Thursday November 28, 2019 @05:42PM (#21660652)
Oh really? As an outsider, that’s super interesting to know. Why can’t system RAM be used in Linux/Windows, or why is it able to be on macOS?
Re:Tensorflow 2.0 AMD Support by rrss on Friday November 29, 2019 @11:36AM (#21665282)
You can do that with OpenCL or CUDA on any platform, it's just ridiculously slow because of the much lower bandwidth so nobody does it. GPUs can access their memory (GDDR or HBM) at 300-800 GB/s, but system memory at only 16 GB/s.
Re:Tensorflow 2.0 AMD Support by shmerl on Thursday November 28, 2019 @08:43PM (#21661360)
Why would anyone care about macOS for Tensorflow though? Even without CUDA, it doesn't support anything portable that targets GPUs either (i.e. OpenCL is in bit rot there, and Vulkan isn't supported to begin with). MacOS is a dead end, and Apple can blame their own extreme lock-in and NIH syndrome for it.
Re:Tensorflow 2.0 AMD Support by rarecoil on Thursday November 28, 2019 @08:00PM (#21661166)
I don't know of a way to get any TensorFlow GPU implementation to work on macOS. I know that Keras runs with PlaidML+Metal on mac, but not TF.

`tensorflow-rocm` assumes Linux/ROCm, and the ROC (Radeon Open Compute) stuff builds Linux kernel extensions (rocm-dkms).

Re:Tensorflow 2.0 AMD Support by losvedir on Thursday November 28, 2019 @09:53AM (#21657721)
How do you mean Google is pushing Swift for Tensorflow? I knew the fast.ai folks were working on it, but I didn't know Google was pushing it at all.
Re:Tensorflow 2.0 AMD Support by dagmx on Thursday November 28, 2019 @10:17AM (#21657867)
Google is the company developing it with Chris Latner in charge
Re:Tensorflow 2.0 AMD Support by pjmlp on Thursday November 28, 2019 @10:49AM (#21658116)
And thanks to it they decided on a language that is barely workable outside Apple's eco-system, instead of more mature alternatives.

I don't have big hopes for the project.

Re:Tensorflow 2.0 AMD Support by girvo on Thursday November 28, 2019 @05:43PM (#21660661)
I’m definitely not disagreeing with you, but the Linux story isn’t completely horrible mostly due to IBM’s work. Though I’ve no idea if they’ve kept pace with the macOS version and tooling, it’s been a couple years since I looked. A shame, I quite like it as a language.
Re:Tensorflow 2.0 AMD Support by pjmlp on Friday November 29, 2019 @01:52AM (#21662603)
IBM seems to have lost interest.

Kitura last commit was on June 2018.

And even with all that support, it is not possible to write OS independent Swift without kludges (import darwin vs glibc, really?).

Also many enterprises impose Windows desktops, so none of their dev divisions would even consider Tensorflows's Swift variant, regardless how well it would be doing on GNU/Linux.