Claude creator Anthropic is considering designing its own chips as advanced AI systems cause a shortage, sources told Reuters.
Anthropic continues to grab the headlines this week, as it fights the US administration in the courts and the power of its unreleased Claude Mythos model strikes fear into the hearts of much of the industry, given its ability to exploit security vulnerabilities.
Now Reuters is citing sources that say Anthropic is looking closely at the possibility of building its own chips, amid industry concerns that the supply of sophisticated chips required for new AI systems from itself and its competitors may not keep pace. Rivals Meta and OpenAI already have such projects underway.
Earlier this week, Anthropic announced a new expanded agreement that will allow it to tap 3.5GW of Google’s tensor processing unit (TPU) capacity from Broadcom.
In a regulatory filing on 6 April, Broadcom said that Anthropic’s consumption of TPU capacity is dependent on its continued commercial success. The multi-gigawatt capacity is expected to come online in 2027.
