Anthropic Confirms In-House AI Chip Development for Claude

ME AI News, according to monitoring by Beating, Anthropic has publicly confirmed for the first time today that it is building an internal chip team to design custom AI chips for Claude. The company will adopt a hardware-and-model co-design approach to enable Claude to “run faster and more efficiently at the scale required by customers,” and has explicitly stated it will employ a multi-chip strategy—hardware from AWS, Google, NVIDIA, and AMD will remain central to its scaling efforts. Previously, Reuters reported that Anthropic was exploring the possibility of developing its own chips, and The Information revealed last month that Anthropic had been negotiating potential manufacturing partnerships with Samsung Electronics. The latest job postings further clarify the details of the plan—Anthropic is recruiting engineers with extensive expertise in semiconductor design and verification to build its custom chip team, offering salaries between $320,000 and $485,000, and requiring candidates to have “direct personal contributions” to the successful tape-out and delivery of semiconductor designs. The job posting specifically emphasizes that “this is a role for individuals who have successfully taped out chips, hold a pragmatic attitude toward timelines, and can make significant decisions with support from a still relatively small organization.” Anthropic has become the latest AI lab to enter the custom chip space—OpenAI launched its own custom chip, Jalapeño, developed in partnership with Broadcom, in June this year. Anthropic’s decision to launch its custom chip initiative comes at a time of surging demand for Claude. Developing in-house chips is seen as a key strategic path for leading AI companies to reduce reliance on external suppliers and achieve deep hardware-software co-optimization amid rising compute costs and constrained capacity from external vendors. Its multi-chip strategy also indicates that Anthropic will not fully脱离 existing chip suppliers in the short term. Click the link below to join Beating’s Feishu AI News Channel for 24/7 real-time monitoring of global AI trends and news. (Source: BlockBeats)