According to reports, Anthropic engineers discovered that using Claude to design chips takes significantly less time than traditional workflows, with comparisons involving chip design processes from companies such as NVIDIA and AMD. This finding was disclosed during a recent investor meeting where the company announced it had formed a 15-person chip design team to work on custom chip engineering and design using its own Claude model.
Insiders stated that Anthropic engineers compared the design time assisted by Claude with their estimates of traditional workflows from large chip design firms like NVIDIA and AMD, concluding that Claude can "significantly shorten" the time required for chip design. The report did not disclose specific shortening ratios or task types, but the comparison benchmark "against traditional processes estimated by NVIDIA and AMD" is noteworthy, indicating that Anthropic's internal reference is based on the two most mature chip design pipelines globally.
Anthropic's chip design team significantly enhances efficiency
This 15-person chip design team was not formed temporarily. In August of this year, Anthropic announced the establishment of an internal team to design custom chips for training and running the Claude AI model. According to four insiders, Anthropic is preparing to launch an IPO later this month, with a valuation potentially reaching $1.5 trillion, and Claude's efficiency in chip design is one of the capabilities the company aims to showcase during the investor meeting. Presenting the ability to "design AI chips with AI" on the eve of the IPO clearly communicates the narrative: Claude is not only a source of revenue but also positioned as a tool to accelerate the construction of its own computing infrastructure.
Anthropic is not the only company validating Claude's efficiency in chip development. In August of this year, Samsung Electronics confirmed the introduction of Claude, with its System LSI division completing a custom SoC verification task originally expected to take over a month in just two days, achieving an efficiency improvement of about 15 times; the task was carried out entirely by an engineer who had been with the company for only two years using Claude Code. Samsung's case provides an independent reference from a leading chip manufacturer, while Anthropic's own findings comparing "against traditional processes from NVIDIA and AMD" extend this efficiency narrative from the verification phase to the earlier custom chip design scenario.
As of now, Anthropic has not disclosed the specific aspects of chip design where Claude is involved or any quantifiable time savings data; further details are awaited from official sources.
Disclosure: This article contains links to affiliate products; if you purchase through these links, TechRitual may earn a commission, but it does not affect product evaluations and recommendations. For more details, please refer to our privacy policy.

