News for hackers. Stuff that matters.
Re:A new GPU backend for the TVM stack by muli on Monday October 30, 2017 @01:29PM (#15587017)
Some quick calculations about TFLOPS per dollar on GEMM.

Since NVIDIA's volta consumer card is not out yet, I used Titan Xp as the reference card. I grabbed prices from wikipedia, and assume TVM reaches 64% peak perf on Vega and 90% peak perf on Titan Xp:

Radeon RX Vega 64: 12.6TFLOPS * 65% / $499 = 0.01638 TFLOPS/$ Pascal Titan Xp: 12TFLOPS * 90% / $1200 = 0.009 TFLOPS/$

So Vega outperforms a lot here.

Re:A new GPU backend for the TVM stack by jamilbk on Monday October 30, 2017 @02:27PM (#15587590)
Another important factor to consider is Vega supports double speed FP16 operations[1] and some ML frameworks are already beginning to optimize for that[2], so that's almost 24 TFLOPS of training compute for ~ $400 USD on the RX Vega 56.

[1]: https://www.anandtech.com/show/11717/the-amd-radeon-rx-vega-... [2]: https://github.com/plaidml/plaidml/issues/29

Re:A new GPU backend for the TVM stack by dharma1 on Monday October 30, 2017 @02:25PM (#15587577)
Another thing to note is that you can in theory train with half precision without sacrificing much accuracy. There isn't that much support for it yet, for AMD anyway that I've seen, but Vega 64 has 2x speed half precision (25 TFLOPS) on paper.

Out of Nvidia cards only Tesla and Volta have 2x (or more) speed mixed precision ops (in Voltas case actually much faster because of the Tensor Cores) but they are in a very different price bracket.

Re:A new GPU backend for the TVM stack by dragontamer on Monday October 30, 2017 @01:46PM (#15587210)
I'd assume that Video RAM is important though. The Titan XP has 12GB of RAM, while the $500 Vega 64 only has 8GB of RAM.

A better comparison is probably Vega Frontier with 16GB of RAM for $1000. If you're doing heavy compute, you're probably gonna need a ton of RAM to go with it.

Re:A new GPU backend for the TVM stack by microcolonel on Monday October 30, 2017 @03:01PM (#15587898)
Vega has half-precision floats as well though, which (with a seemingly negligible loss in precision) in combination with their HBCC (transparent main memory DMA) should more than make up for the lack of memory on the Vega 64, and in the case of Frontier Edition, all the more so.
Re:A new GPU backend for the TVM stack by dharma1 on Monday October 30, 2017 @02:36PM (#15587680)
Vega is meant to have someting called HBCC - that effectively allows you to use system RAM or even SSD as GPU memory for VERY large datasets. I have no idea if it would be fast enough for deep learning workloads over PCIe

https://www.pcper.com/news/General-Tech/AMDs-HBCC-you-and-me

Re:A new GPU backend for the TVM stack by jamilbk on Monday October 30, 2017 @02:34PM (#15587654)
That's a fair assumption I think. It will be interesting to see whether Vega's high-bandwidth cache controller (HBCC) will help nullify this difference if implemented in ML frameworks.
Re:A new GPU backend for the TVM stack by TomV1971 on Monday October 30, 2017 @04:41PM (#15588690)
AMD positions the $1000 Vega FE against the $1200 Titan Xp.

The $700 1080 Ti has almost the same performance as the Titan Xp, so why not compare those?

It changes conclusions considerably...

Re:A new GPU backend for the TVM stack by dragontamer on Monday October 30, 2017 @04:57PM (#15588808)
Depends on the situation, which is why I think Vega Frontier Edition ($1000) is the apt comparison against the Titan XP ($1200).

The 1080 Ti has less RAM, and has NO support for 16-bit packed arithmetic. Ultimately, the 1080 TI is a graphics card designed to dominate video games, and NVidia cuts out other features that gamers don't care about.

With regards to the "Compute" sector, its Vega Frontier ($1000) vs Titan XP ($1200) at the low-end at least. NVidia Tesla chips ($4000 to $7000) constitute the higher-end.

Re:A new GPU backend for the TVM stack by 15155 on Monday October 30, 2017 @06:35PM (#15589592)
> and NVidia cuts out other features that gamers don't care about.

Specifically: substantially better drivers for compute.