Amazon Announces Supercomputer, New Server in Push To Take on Nvidia [Video]

Amazon (AMZN) is building an artificial intelligence (AI) supercomputer made of its in-house Trainium chips and also a new server that it said will be used by AI startup Anthropic.

The “Ultracluster” announced by the tech giant’s Amazon Web Services (AWS) cloud computing arm was designed by Amazon’s chip lab in Austin, Texas, The Wall Street Journal reported. Amazon last month invested an additional $4 billion in Anthropic, whose Claude chatbot is a rival to OpenAI’s ChatGPT.

“Together with Anthropic, AWS is building an EC2 UltraCluster of Trn2 UltraServers—named Project Rainier—containing hundreds of thousands of Trainium2 chips and more than 5x the number of exaflops used to train their current generation of leading AI models,” Amazon said. 

An exaflop is a measure of a supercomputer’s performance.

According to the Journal, the new chips and servers highlight AWS’s commitment to the homegrown Trainium it is positioning as a “viable alternative” to the graphics processing …