NVIDIA HGX-2 HPC/AI Server Platform Offers 16 x V100 GPUs, 2 PFLOPS of Tensor Cores
Written by Michael Larabel in NVIDIA on 30 May 2018 at 05:36 AM EDT. 14 Comments
The HGX-2 is an impressive beast, but will cost an incredible amount too.

NVIDIA has introduced the HGX-2 as their newest and most advanced cloud server platform intended for HPC and API workloads. The HGX-2 offers an incredible sixteen V100 (Volta) GPUs connected via NVLink yielding 2 PetaFLOPS worth of tensor cores, 512GB of video memory, and 2,400GB/s of bisection bandwidth. This more than doubles the performance capabilities of the previous HGX-1 server platform.

NVIDIA claims that the HGX-2 can replace up to 300 CPU-only servers for some machine/deep learning scenarios. Those wanting to learn more about the NVIDIA HGX-2 can do so at NVIDIA.com.
Related News
About The Author
Author picture

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter or contacted via MichaelLarabel.com.

Popular News This Week