NVIDIA hops onto next generation of accelerated computing architecture

After Pascal, Volta and Ampere comes Hopper, the next generation of NVIDIA’s accelerated computing architecture. Announced by Founder and CEO Jensen Huang at his GTC keynote this morning, Hopper is expected to deliver a massive performance leap over its predecessor to power the next wave of AI data centres.

It is named after Grace Hopper, a pioneering US computer scientist who devised the theory of machine-independent programming languages and whose Flow-Matic programming language was used to create the highly popular common business oriented language (COBOL).

“Data centres are becoming AI factories — processing and refining mountains of data to produce intelligence. NVIDIA H100 is the engine of the world’s AI infrastructure that enterprises use to accelerate their AI-driven businesses,” said Huang.

NVIDIA’s first Hopper-based GPU is the NVIDIA H100, which is packed with 80 billion transistors and features a revolutionary Transformer Engine and a highly scalable NVIDIA NVLink interconnect for advancing gigantic AI language models, deep recommender systems, genomics, and complex digital twins.

Its eight H100 GPUs deliver 32 petaflops of AI performance at new FP8 precision. They ares connected by fourth-generation NVLink, which provides 900GB/s connectivity, 1.5x more than the prior generation. NVSwitch enables all eight GPUs to connect over NVLink.

The NVIDIA H100 GPU is the first accelerator with confidential computing capabilities to protect AI models and customer data while they are being processed. This is particularly useful for privacy-sensitive industries such as healthcare and financial services, as well as on shared cloud infrastructures.

NVIDIA H100 can be deployed in every type of data centre — on-premises, cloud, hybrid-cloud and edge. It is expected to be available worldwide from Q3 from the world’s leading cloud service providers and computer makers, as well as directly from NVIDIA.

A wide range of servers with H100 accelerators are expected from systems manufacturers such as Atos, BOXX Technologies, Cisco, Dell Technologies, Fujitsu, Gigabyte, H3C, Hewlett Packard Enterprise, Inspur, Lenovo, Nettrix, and Supermicro.