Graphics Cards

Nvidia ampere would have 50% better performance compared to turing

Table of contents:

Anonim

The state-of-the-art Nvidia "Ampere" GPU architecture has been one of the company's best-kept secrets, but it seems that its partners in the server market are standing up for them. According to The Next Platform (via Videocardz ), Indiana University will roll out its latest Big Red 200 supercomputer, which will reportedly use the green company's new GPU, with huge performance improvements over the current generation Turing..

Nvidia Ampere will be present in the Big Red 200 supercomputer

According to Brad Wheeler, vice president of information technology and chief information officer at Indiana University, the Big Red 200 supercomputer marks the university's 200th anniversary. The Big Red 200 'supercomputer' is claimed to be compatible with conventional HPC workloads as well as AI-specific workloads. The supercomputer will be designed in two phases, the first part, which is already in operation, consists of 672 dual-socket nodes that feature AMD's EPYC 7742 processors, which have 64 cores and 128 threads, for a total of 86, 016 cores and 172, 032 threads available.

The second phase of the supercomputer is the most interesting. Once it goes live this summer, it would not only put more EPYC 7742 processors to work, it would also have Nvidia's next-generation GPUs. What's interesting is that Big Red 200 was originally slated to incorporate the Tesla V100 GPUs based on the Volta architecture, but according to Wheeler, they planned to wait a bit longer for Nvidia's next-gen parts, and that's when the idea for developing the system came up. in two phases.

Initially, the Big Red 200 peak performance was around 5.9 petaflops, but now it is expected to provide 8 petaflops of computing power. Each node of the Big Red 200 supercomputer will be equipped with one or two new generation Nvidia GPUs.

Visit our guide on the best graphics cards on the market

It is also mentioned that Big Red 200 got 2 additional petaflops of performance, although it uses a smaller number of GPUs than the design based on the Volta V100. The reason for opting for a smaller number of next-generation GPUs is simply because they offer 70-75% more performance than existing parts and that's why we compare it to Volta based Tesla V100 GPUs as there are no Tesla based GPUs in Turing's GPU architecture, apart from the Tesla T4.

If the 70-75% performance boost figures are true on the Big Red 200, then we can definitely see about 50% performance boost on Ampere or even more on consumer variants hitting stores throughout the year. of this year. We will keep you informed.

Wccftech font

Graphics Cards

Editor's choice

Back to top button