Project Rainier, under construction by Amazon Web Services, is set to be a massive supercomputing cluster featuring hundreds of thousands of accelerators. It will enable Anthropic to perform model training on an unprecedented scale across multiple US sites. A notable site in Indiana will include thirty datacenters consuming over 2.2 gigawatts of power. The project emphasizes the deployment of Amazon's own Annapurna AI silicon rather than traditional GPUs, and it reflects Amazon's substantial investment in Anthropic, aimed at advancing competitive AI capabilities.
This is the first time we are building such a large-scale training cluster that will allow a customer, in this case Anthropic, to train a single model across all of that infrastructure."
The scale is really unprecedented.
However, we're told that Anthropic has already managed to get their hands on a sliver of the system's compute.
Amazon is among Anthropic's biggest backers, having already invested $8 billion in the OpenAI rival.
Collection
[
|
...
]