NVIDIA declares DGX GH200, its new class of AI supercomputers: All you’ll want to know

US multinational tech large NVIDIA has introduced a brand new class of large-memory synthetic intelligence (AI) supercomputers, which it says shall be referred to as DGX GH200.

These supercomputers will ‘enable the development of giant, next-generation models for generative AI language applications, recommender systems, and data analytics workloads,’ NVIDIA mentioned in a assertion on its official weblog on May 28.

“Generative AI, large language models and recommender systems are the digital engines of the modern economy. DGX GH200 supercomputers integrate NVIDIA’s most-advanced accelerated computing and networking technologies to expand the frontier of AI,” the assertion quoted Jensen Huang, the corporate’s founder-CEO, as saying.

Here is all you’ll want to learn about NVIDIA’s DGX GH200 supercomputer:

(1.) The machine is powered by GH200 Grace Hopper Superchips and NVLink Switch System, each in-house merchandise; it additionally has an enormous shared reminiscence house.

(2.) The reminiscence house combines the superchips, permitting the latter to carry out as a single graphic processing unit (GPU). This, in flip, supplies almost 500 occasions extra reminiscence than 2020’s earlier technology NVIDIA DGX A100 supercomputers.

(3.) DGX GH200, which is the primary supercomputer to mix superchips with the Switch System, has an structure that gives 48 time extra bandwidth than that within the earlier technology.

(4.) Google Cloud, Meta and Microsoft are among the many first shoppers which can be anticipated to get entry to DGX GH200, and can discover its capabilities for generative AI workloads.

(5.) NVIDIA plans to launch these machines by the tip of the 12 months.