Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Nvidia DGX-1 Deep Learning Supercomputer in a Box

nvidia.com

Post reply on HN
Print view View original

81–90 of 106 posts

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#81

Post by trsohmers »

mortenjorck wrote: post #68

Just for some perspective, a little over 10 years ago, this $130k turnkey installation would sit at #1 in TOP500, easily beating out hundred-million-dollar initiatives like NEC's Earth Simulator and IBM's BlueGene/L: http://www.top500.org/lists/2005/06/ (170 TFLOPS vs. 137 TFLOPS) At the other end, even a single GTX 960 would make it onto the list, placing in the 200s.

The 170 TFLOPs number that NVIDIA gives out is for FP16, while the Top 10 list gives its number for for FP64. The P100 that makes up this NVIDIA box gives about 5.3TFLOPs per card, or a total of 42.4TFLOPs for the whole box.

Sure, you can say that deep learning doesn't need FP64, but it is REALLY unfair to compare this to anything on the TOP500 list, especially when you consider the fact that this is not balanced in terms of memory size or bandwidth (in relation to the number of FLOPs) when you compare it to any real supercomputer class system.

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#82

Post by jlebar »

phelm wrote: post #18

I am looking forward to OpenCL catching up with CUDA in maturity and adoption, so that NVidia's monopoly in Silicon for deep learning will come to an end.

clang now has a mostly-working CUDA frontend (disclaimer, I work on it). And it has an AMD GPU backend (whether this is in a good state I don't know). I don't expect that putting these pieces together would be a huge project.

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#84

Post by KKKKkkkk1 »

mortenjorck wrote: post #68

Just for some perspective, a little over 10 years ago, this $130k turnkey installation would sit at #1 in TOP500, easily beating out hundred-million-dollar initiatives like NEC's Earth Simulator and IBM's BlueGene/L: http://www.top500.org/lists/2005/06/ (170 TFLOPS vs. 137 TFLOPS) At the other end, even a single GTX 960 would make it onto the list, placing in the 200s.

Theoretical peak flops rate is useless for indicating performance nowadays. There are new benchmarks that take memory and network performance into account such as HPCG and HPGMG. On these benchmarks, throughput-oriented machines such as the ones Nvidia sells do not look good at all.

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#85

Post by doyoulikeworms »

trsohmers wrote: post #81
mortenjorck wrote: post #68

Just for some perspective, a little over 10 years ago, this $130k turnkey installation would sit at #1 in TOP500, easily beating out hundred-million-dollar initiatives like NEC's Earth Simulator and IBM's BlueGene/L: http://www.top500.org/lists/2005/06/ (170 TFLOPS vs. 137 TFLOPS) At the other end, even a single GTX 960 would make it onto the list, placing in the 200s.

The 170 TFLOPs number that NVIDIA gives out is for FP16, while the Top 10 list gives its number for for FP64. The P100 that makes up this NVIDIA box gives about 5.3TFLOPs per card, or a total of 42.4TFLOPs for the whole box. Sure, you can say that deep learning doesn't need FP64, but it is REALLY unfair to compare this to anything on the TOP500 list, especially when you consider the fact that this is not balanced in…

Out of curiosity, what are some problems/solutions that require FP64?

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#87

Post by zevets »

doyoulikeworms wrote: post #85
trsohmers wrote: post #81

Earlier quoted context omitted.

The 170 TFLOPs number that NVIDIA gives out is for FP16, while the Top 10 list gives its number for for FP64. The P100 that makes up this NVIDIA box gives about 5.3TFLOPs per card, or a total of 42.4TFLOPs for the whole box. Sure, you can say that deep learning doesn't need FP64, but it is REALLY unfair to compare this to anything on the TOP500 list, especially when you consider the fact that this is not balanced in…

Out of curiosity, what are some problems/solutions that require FP64?

Any finite element/volume problem. Anything integrated.

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#88

Post by Fomite »

olympus wrote: post #39
badminton1 wrote: post #21

Costs $129,000 and needs 3.2 kilowatts to run.

To me, $129k isn't surprising since it is only going to be bought by researchers with big budgets. Small-timers will still build 3x GTX980 systems for under $5k. 3.2 KILOwatts sounded insane to me, but I suppose you'll have your own server rack to put it in if you can afford to buy one of these.

"To me, $129k isn't surprising since it is only going to be bought by researchers with big budgets" Yeah, this is essentially "Big chunk of a computational researcher's startup budget" or an infrastructure grant.

Re: The Nvidia DGX-1 Deep Learning Supercomputer in a Box

#90

Post by pjmlp »

nl wrote: post #38
phelm wrote: post #18

I am looking forward to OpenCL catching up with CUDA in maturity and adoption, so that NVidia's monopoly in Silicon for deep learning will come to an end.

The problem is that no one[1] uses OpenCL because the performance isn't there. There is little sign of that changing, too.

Another reasons are language support and tooling.

CUDA had Fortran and C++ since day one and thanks to PTX was quite easy to add support for other languages.

Whereas OpenCL was stuck on "C only" model from Khronos, which forced everyone to use C or generate C code and be constrained to the device drivers.

This has been seen as such a big issue that SPIR and C++ SPIR got introduced with OpenCL 2.0.

Another very important one is debugging support. Last time I checked no one had visual tooling at the same level as NVidia's one.

Post reply on HN