Anthropic is making its custom-chip ambitions official.
The Claude developer is assembling an in-house silicon team to work on custom chips, according to Business Insider and a company job listing seeking engineers with experience completing and shipping semiconductor designs.
The move places Anthropic among a growing group of AI companies trying to reduce infrastructure constraints by designing hardware and software together. The larger question is how much control Anthropic can gain while still relying heavily on cloud and manufacturing partners.
What Anthropic announced
According to Business Insider, a job listing from Anthropic is calling for candidates with experience in different domains of chip design and verification. The job description also notes that applicants must demonstrate “direct personal contribution” to the completion and shipment of semiconductor designs.
Those requirements indicate that Anthropic is seeking engineers with experience taking chips beyond the research stage and into production. The role carries a salary ranging from $320,000 to $485,000, underscoring the premium the company is willing to pay for experienced chip engineers.
In a statement reported by Business Insider, the company said that the new team forms part of its broader “multi-chip” strategy and will complement — not replace — its existing use of hardware from Amazon Web Services(AWS), Google, Nvidia, and AMD.
Anthropic did not disclose when its first custom chip could be ready or whether it intends to manufacture the processors itself.
The effort follows a June report from The Information, which said Anthropic was in talks with Samsung Electronics over plans to help build custom AI chips. However, the company had not publicly confirmed the effort at the time.
More must-read AI coverage
- SS&C Intralinks DealCentre AI vs. Datasite: Which platform is built for the future of dealmaking?
- SS&C Intralinks FundCentre AI vs. Juniper Square: Which platform better supports modern private markets fund managers?
- Why Data, Not Models, Determines AI Success
- The Rise of the AI-Native Factory: How Physical AI Is Transforming Manufacturing
Why Anthropic is building its own chips
Building an in-house chip team gives the company greater control over access to the hardware powering its AI models. That has become increasingly important given the current state of affairs with AI infrastructure.
Designing chips around Claude’s workloads also allows Anthropic to optimize its hardware and software together, potentially improving performance and efficiency while reducing the long-term costs of training and running increasingly large AI models.
The strategy is not without precedent. Apple’s transition to its in-house Apple Silicon chips transformed the performance and power efficiency of Macs by designing hardware specifically for its software. While Anthropic’s goals differ, the same principle applies: hardware built for a specific workload can often deliver better results than hardware designed to serve everyone.
Why this matters beyond Anthropic
Anthropic’s announcement is another sign that AI companies are no longer competing only on the quality of their models. Increasingly, the race is extending to the infrastructure beneath them, as developers seek to secure computing capacity, improve efficiency, and reduce dependence on a limited supply of AI chips.
For users, the benefits may not be immediate, but they could become noticeable over time.
Recently, Moonshot AI temporarily paused registrations for its Kimi K3 after demand surged past available infrastructure. That is the kind of issue Anthropic is trying to avoid for users.
AI services powered by purpose-built hardware can eventually become faster and less expensive to operate. If many AI companies adopt Anthropic’s approach, that could potentially allow them to deploy larger and more capable models at a lower cost and at scale.
Other News: Google’s August Android 17 update fixes gaming crashes, GPU performance issues, and unresponsive touchscreens on Pixel 10 devices while leaving several older Pixel models off the stable rollout.