Anthropic Appoints Former Google Chip Head to Lead New Hardware Initiative

Yahoo Finance ·

Artificial intelligence startup Anthropic has recruited former Google executive Amir Salek to its compute team as the company initiates an in-house semiconductor program, Bloomberg reported. Salek, who previously spearheaded Alphabet Inc.'s custom silicon strategy and successfully delivered seven generations of Tensor Processing Units, will report directly to Anthropic compute leader James Bradbury. Most recently a senior managing director at Cerberus Capital Management following his 2022 departure from Google, Salek brings extensive enterprise chip scaling experience to the San Francisco-based lab. This executive hire highlights Anthropic's strategic shift toward proprietary hardware architecture to tailor designs to operational needs and bypass supply chain bottlenecks, even as it continues procuring chips from Nvidia Corp., Amazon.com Inc., and Google. The move mirrors broader industry trends, with rival OpenAI similarly partnering with Broadcom Inc. to develop its custom Jalapeno chip. To bridge immediate infrastructure needs, Anthropic has committed $250 million to UK chip firm Fractile while securing expanded data center capacity through agreements with Riot Platforms Inc. and Volta Infra Holdings Ltd.

AI 시장 분석

Anthropic has hired former Google custom silicon head Amir Salek, officially launching its in-house semiconductor development. This is a strategic move by AI labs to reduce dependence on external suppliers like NVIDIA and cut long-term computing costs. Investors must closely monitor the long-term impact of AI companies internalizing their own chips on existing semiconductor supply chains and hardware ecosystems.

상승 영향

하락 영향

DYAX 전담 분석

Anthropic's entry into in-house semiconductor development shows that the trend of hardware internalization is accelerating in the AI infrastructure market. By recruiting a key talent who successfully led Google's 7th generation TPU, Anthropic has laid the foundation to reduce chip procurement costs and alleviate supply chain bottlenecks in the long run.

The bullish scenario is that the success of proprietary chips maximizes AI model operation efficiency, while the bearish scenario is a delay in securing effectiveness relative to massive development cost investments. Going forward, the development speed of proprietary silicon by AI labs and the earnings trends of existing chip suppliers should be monitored as key indicators.

AI가 생성한 분석으로 투자 자문이 아닙니다.

DYAX Investor Sentiment

Bullish (Long) 59% · Bearish (Short) 41%

311 participants

Related News

원문 보기 — Yahoo Finance