A naukri.com initiative
Insider
2w
157
Image Credit: Insider
Silicon and supercomputers will define the next AI era. AWS just made a big bet on both.
- Amazon Web Services unveiled a new AI chip Trainium3 and announced a new supercomputer that will be built with its own chips to serve its AI ambitions.
- This move shows that Amazon is reducing its reliance on Nvidia for AI chips and is ready to control silicon and supercomputers of their own.
- Amazon's Trainium2 chip is generally available now and it offers 30-40% better price performance than the current generation of servers with Nvidia GPUs.
- Trainium3, which AWS previewed ahead of a late 2025 release, has been billed as a "next-generation AI training chip." Servers loaded with Trainium3 chips offer four times greater performance than those packed with Trainium2 chips.
- Google and OpenAI are also exploring custom, in-house chip designs of their own to reduce their dependence on Nvidia.
- AWS shared that it was working with Anthropic to build an UltraCluster of servers that form the basis of a supercomputer it has named Project Rainier.
- When completed, Project Rainier will be the world's largest AI compute cluster reported to date available for Anthropic to build and deploy their future models on.
- AWS also acknowledged that companies serious about building their own AI models will need access to highly specialized computing that can handle a new era of AI.
- Nvidia is still responsible for 99% of the workloads for training AI models today, but AWS believes Trainium can carve out a good niche for itself.
- Other big tech companies are also designing their own AI chips and supercomputers, including Microsoft and Elon Musk's company xAI building a supercomputer using 100,000 Nvidia GPUs in Memphis this year.
Read Full Article
9 Likes
For uninterrupted reading, download the app