Arthur J. Villasanta – Fourth Estate Contributor
Santa Clara, CA, United States (4E) – NVIDIA Corporation claims to have developed the “fastest single computer humanity has ever created,” which is the HGX-2 GPU server built to accelerate artificial intelligence (AI) and high performance computing (HPC).
NVIDIA HGX-2 is designed for multi-precision computing. This type of computing combines the power of high-precision scientific computing using FP64 and FP32 with the speed of lower-precision AI computing with FP16 and Int8 to provide a single flexible and powerful platform to solve these massive challenges.
Accelerated by 16 NVIDIA Tesla V100 GPUs and NVIDIA NVSwitch, HGX-2 has what the company says is “the unprecedented compute power, bandwidth, and memory topology to train these models faster and more efficiently.” The 16 Tesla V100 GPUs work as a single unified 2 petaFLOP accelerator with half a terabyte (TB) of total GPU memory, allowing it to handle the most computationally intensive workloads and enable “the world’s largest GPU.”
With a half-terabyte of high-bandwidth memory (HBM), NVIDIA HGX-2 is said to be capable of replacing as many as 300 CPU-driven servers on its own, offering a much more powerful and compact computing solution for data-driven tasks.
NVIDIA CEO Jensen Huang claims the 16 Tesla V100 graphics chips can “talk to every one of the GPUs simultaneously at a bandwidth of 300 GB/s, 10 times PCI Express, so everyone can talk to each other all at the same time.” This reveal of the NVIDIA HGX-2 was announced at NVIDIA’s own GPU Technology Conference in Taiwan.
“With the HGX-2 server building block, we’ll be able to quickly develop new systems that can meet the growing needs of our customers who demand the highest performance at scale,” said Paul Ju, Lenovo’s vice president and general manager.
HPC applications require strong server nodes with the computing power to perform a massive number of calculations per second. Increasing the compute density of each node dramatically reduces the number of servers required, resulting in huge savings in cost, power, and space consumed in the data center.
Article – All Rights Reserved.
Provided by FeedSyndicate