{"id":2145,"date":"2022-03-23T15:13:13","date_gmt":"2022-03-23T15:13:13","guid":{"rendered":"https:\/\/scienceandnerds.com\/2022\/03\/23\/nvidia-reveals-h100-gpu-for-ai-and-teases-worlds-fastest-ai-supercomputer\/"},"modified":"2022-03-23T15:13:14","modified_gmt":"2022-03-23T15:13:14","slug":"nvidia-reveals-h100-gpu-for-ai-and-teases-worlds-fastest-ai-supercomputer","status":"publish","type":"post","link":"https:\/\/scienceandnerds.com\/2022\/03\/23\/nvidia-reveals-h100-gpu-for-ai-and-teases-worlds-fastest-ai-supercomputer\/","title":{"rendered":"Nvidia reveals H100 GPU for AI and teases \u2018world\u2019s fastest AI supercomputer\u2019"},"content":{"rendered":"

Source: https:\/\/www.theverge.com\/2022\/3\/22\/22989182\/nvidia-ai-hopper-architecture-h100-gpu-eos-supercomputer<\/a>
\n
<\/br><\/code><\/p>\n

\n

Nvidia has announced a slew of AI-focused enterprise products at its annual GTC conference. They include details<\/a> of its new silicon architecture, Hopper; the first datacenter GPU built using that architecture, the H100<\/a>; a new Grace CPU \u201csuperchip\u201d<\/a>; and vague plans to build what the company claims will be the world\u2019s fastest AI supercomputer, named Eos<\/a>.<\/p>\n

Nvidia has benefited hugely from the AI boom of the last decade, with its GPUs proving a perfect match for popular, data-intensive deep learning methods. As the AI sector\u2019s demand for data compute grows, says Nvidia, it wants to provide more firepower.<\/p>\n

In particular, the company stressed the popularity of a type of machine learning system known as a Transformer. This method has been incredibly fruitful, powering everything from language models like OpenAI\u2019s GPT-3 to medical systems like DeepMind\u2019s AlphaFold. Such models have increased exponentially in size over the space of a few years. When OpenAI launched GPT-2 in 2019, for example, it contained 1.5 billion parameters (or connections). When Google trained a similar model just two years later, it used 1.6 trillion<\/em> parameters<\/a>.<\/p>\n

\n