Nvidia reveals why it chose rival AMD over Intel for its deep learning system

[midian182]

In context: Like Sony vs. Microsoft, Intel vs. Qualcomm, and Apple vs. everyone, Nvidia vs. AMD is one of the tech industry's big rivalries. So, it came as a surprise when team green chose its main competitor to provide the server processors for its new DGX A100 deep learning system, rather than using Intel’s Xeon platform. Now, the company has revealed the reason behind its decision.

In Nvidia’s first two DGX systems, Intel’s Xeon CPUs were the preferred processor, but the company dropped them in the DGX A100 for two of AMD's 64-core, Zen 2-based Epyc 7742 CPUs. The system, which uses the new, Ampere-based A100 GPUs, boasts 5 petaflops of AI compute performance and 320 GB of GPU memory with 12.4 TB per second of bandwidth.

Speaking to CRN, Nvidia’s Vice President and General Manager of DGX Systems, Charlie Boyle, said the decision came down to the extra features and performance offered by the Epyc processors. "To keep the GPUs in our system supplied with data, we needed a fast CPU with as many cores and PCI lanes as possible. The AMD CPUs we use have 64 cores each, lots of PCI lanes, and support PCIe Gen4," he explained.

In addition to having eight more cores than the Xeon Platinum 9282, Epyc 7742 also supports eight-channel memory, whereas Intel’s Xeon Scalable processors support just six memory channels. AMD’s offering is also a lot cheaper—$6,950 vs around $25,000—and has more cache and a lower TDP.

PCIe 4.0 support is one of the major factors for choosing Epyc, with Intel’s processors still only supporting PCIe 3.0. It means AMD's CPUs offer 128 lanes and a peak PCIe bandwidth of 512GB/s. "The DGX A100 is the first accelerated system to be all PCIe Gen4, which doubles the bandwidth from PCIe Gen3. All of our IO in the system is Gen4: GPUs, Mellanox CX6 NICs, AMD CPUs, and the NVMe drives we use to stream AI data," Boyle said.

AMD, of course, has the advantage of using the 7nm manufacturing process, though Intel’s 10nm Ice Lake server CPUs, which are expected feature PCIe 4.0 support, arrive later this year.

Permalink to story.

https://www.techspot.com/news/85301-nvidia-reveals-why-chose-rival-amd-over-intel.html

 

[Uncle Al]

Certainly sounds like a wise decision ... we'll see how it turns out in the long run.

 

[grumblguts]

AMD FTW

 

[Irata]

Given their IO and other requirements and the fact that Intel wants to compete with nVidia in the HPC GPU compute market, as well, it was an obvious choice.

Also, customers buy these systems for the GPU power, so why should nVidia not be pragmatic and chose supporting components based on what makes their system run best? That is Epyc now, may be Intel, IBM or ARM in the future....

 

[veLa]

It's a fairly logical decision.

It also goes to show you that despite being competitors, these companies collaborate more often than you'd expect.

 

[Lionvibez]

This was pretty obvious.

 

[gagegfg]

Procesador Intel® Xeon® Platino 9282
Max Memory Channels "12" not "6". (https://ark.intel.com/content/www/u...atinum-9282-processor-77m-cache-2-60-ghz.html)

"pci-e 4.0" / "$$" / "availability"
those are the reasons for NVIDIA

 

[Tams80]

Those are the main reasons, but I wouldn't be surprised if there's a bit of payback at Intel trying to muscle into the GPU industry, especially given their business practices.

Nvidia and AMD have a rivalry, but they've played clean and amicably. Intel have a terrible reputation. It's not surprising they want more than triple for something better than their offering either.

 

[krizby]

DGX-2 price was 400k while DGX-3 cost 200k, it seems like Nvidia was able to make those cost cutting by using AMD EPYC CPU, neat...