Amazon's Bold Move in AI Computing
Amazon Web Services (AWS), the cloud computing arm of Amazon, has announced plans to develop 'Ultracluster,' a groundbreaking AI supercomputer powered by its Trainium chips. The announcement was made at the re:Invent conference, setting the stage for a 2025 launch.
The Ultracluster is set to support AI startup Anthropic and will offer a staggering five times the computing capacity of AWS's current systems. This new system integrates Amazon's advanced networking technology, delivering up to 83.2 petaflops of computing power, which is four times that of Nvidia's typical chip servers.
A New Era of Choice
In addition to the Ultracluster, AWS unveiled a new server called Ultraserver, composed of 64 interconnected chips. "Today, there's really only one choice on the GPU side, and it's just Nvidia," said Matt Garman, CEO of Amazon Web Services. "We think that customers would appreciate having multiple choices."
Comments