← NewsAll
Nvidia announces Vera Rubin AI chip platform with higher performance
Summary
At CES, Nvidia said its Vera Rubin chip platform is in "full production" and will arrive later this year; the company said it can deliver about five times the AI computing of its previous chips.
Content
Nvidia said at the Consumer Electronics Show in Las Vegas that its next-generation Vera Rubin chip platform is in "full production" and will arrive later this year. Jensen Huang, the company's chief executive, described the platform as aimed at speeding up AI tasks such as serving chatbots and other applications. He said Rubin hardware is being tested in Nvidia's labs by AI firms. The announcement comes as Nvidia faces increasing competition from traditional rivals and from large customers developing their own chips.
Key details:
- Nvidia described the Vera Rubin platform as made up of six separate chips, with a flagship server that contains 72 of the company's graphics units and 36 of its new central processors.
- The company said Rubin can deliver about five times the artificial-intelligence computing of its previous chips for serving chatbots and other AI apps.
- Huang demonstrated how Rubin chips can be linked into "pods" with more than 1,000 chips and said they could improve token-generation efficiency by about 10 times.
- Nvidia said the Rubin chips use a proprietary kind of data the company hopes the wider industry will adopt.
- New features highlighted include a layer called "context memory storage" aimed at helping chatbots respond more quickly in long conversations and a new generation of networking switches using co-packaged optics.
- Other announcements included wider release of the Alpamayo software and the training data used for it, and Nvidia's recent acquisition of talent and chip technology from the startup Groq.
Summary:
Nvidia presents the Vera Rubin platform as a step to increase speed and efficiency when serving AI applications, and expects products to arrive later this year. Industry testing is underway in Nvidia labs, and how the platform is adopted and how competitors respond will determine its broader market impact.
