Anthropic tells IPO investors Claude helps design its chips
TL;DR
- Anthropic told IPO investors its 15-person chip design team is already using Claude on custom silicon meant to train and run Claude.
- Engineers claimed Claude cuts chip-design time versus their estimates of workflows at Nvidia and AMD; no plan to sell the design tooling.
- Anthropic said it will keep a multi-chip approach, with hardware from AWS, Google, Nvidia and AMD central to its compute pool.
The Information reported that Anthropic has been telling prospective IPO investors in recent meetings that its new 15-person chip design team is already using Claude for engineering and design work on the custom silicon meant to train and run Claude. There is no announced plan to sell the design tools.
The specific pitch to investors: Claude 'cuts time taken to design chips' compared with the engineers' estimates of workflows at designers such as Nvidia and AMD. A comparative claim against two of the industry's biggest silicon teams, made by a group of fifteen.
Anthropic also told investors it would continue a 'multi-chip approach,' with Nvidia GPUs, AMD GPUs, AWS Trainium and Google TPUs central to its compute pool. The custom-silicon work is additive, not a replacement for the merchant chips already in production. It lands alongside this week's other Anthropic disclosures on our tracker, including the amended S-1A noting Claude may 'imminently' hit a recursive self-improvement threshold.
Originally reported by theinformation.com
Read the original article →Original headline: Anthropic Tells IPO Investors Its 15-Person Chip Design Team Is Already Using Claude