Story image

NVIDIA introduces HGX-2, fusing HPC and AI computing

NVIDIA recently introduced NVIDIA HGX-2, the first unified computing platform for both artificial intelligence and high-performance computing. 

The HGX-2 cloud server platform, with multi-precision computing capabilities, supposedly provides unique flexibility to support the future of computing. 

It allows high-precision calculations using FP64 and FP32 for scientific computing and simulations, while also enabling FP16 and Int8 for AI training and inference. 

This unprecedented versatility meets the requirements of the growing number of applications that combine HPC with AI. 

A number of leading computer makers today shared plans to bring to market systems based on the NVIDIA HGX-2 platform. 

HGX-2-serves as a “building block” for manufacturers to create some of the most advanced systems for HPC and AI. 

It has achieved record AI training speeds of 15,500 images per second on the ResNet-50 training benchmark and can replace up to 300 CPU-only servers. 

It incorporates such breakthrough features as NVIDIA NV Switch interconnect fabric, which seamlessly links 16 NVIDIA Tesla V100 Tensor Core GPUs to work as a single, giant GPU delivering two petaflops of AI performance. 

The first system built using HGX-2 was the recently announced NVIDIA DGX-2. 

HGX-2 comes a year after the launch of the original NVIDIA HGX-1, at Computex 2017. 

The HGX-1 reference architecture won broad adoption among the world’s leading server makers and companies operating massive data centres, including Amazon Web Services, Facebook and Microsoft. 

OEM, ODM Systems Expected Later This Year Four leading server makers, Lenovo, QCT, Supermicro and Wiwynn announced plans to bring their own HGX-2-based systems to market later this year. 

HGX-2 is a part of the larger family of NVIDIA GPU-Accelerated Server Platforms, an ecosystem of qualified server classes addressing a broad array of AI, HPC and accelerated computing workloads with optimal performance. 

Supported by major server manufacturers, the platforms align with the data centre server ecosystem by offering the optimal mix of GPUs, CPUs and interconnects for diverse training (HGX-T2), inference (HGXI2) and supercomputing (SCX) applications. 

Customers can choose a specific server platform to match their accelerated computing workload mix and achieve best-in-class performance.

The silver lining in Australia’s Government cloud strategy
Cloud has been a huge part of the ‘digital transformation’ conversation within Australian government during recent years.
Aerohive achieves ISO/IEC 27001 cloud platform certification
Aerohive is the first cloud-managed networking vendor recognized by a global standard for commitment to information security management systems.
Is Google’s Stadia feasible with today’s data centres?
To get a better idea of the sheer audacity behind Google’s latest move, we spoke to Unitas Global chief technical officer Grant Kirkwood.
Survey: IT pros nostalgic over on-prem data centre visibility
There are significant security and monitoring challenges faced by IT staff responsible for managing public and private cloud deployments.
61% of CIOs believe employees leak data maliciously
Egress conducted a survey to examine the root causes of employee-driven data breaches, their frequency, and impact.
VMware allures APJ channel veteran to take the reins
Balasingam will take on the role of vice president for VMware’s partner business in Asia Pacific and Japan (APJ).
Security top priority for Filipinos when choosing a bank - Unisys
Filipinos have greatest appetite in Asia Pacific to use biometrics to access banking services
Opinion: Modular data centers mitigate colocation construction risks
Schneider's Matthew Tavares believes modular data centers are key for colocation providers seeking a competitive advantage with rapid deployment.