Theme:
Light Dark Auto
GeneralPoliticsBusinessTechnologyEnvironmentSportsEntertainment
TECHNOLOGY
Neutral Sentiment

Anthropic Hires Google TPU Pioneer to Build Custom AI Chips

PUBLISHED Aug 21, 2026, 7:52 PM ET

Read, Watch or Listen

Media Bias Meter
Sources: 12
Center 100%
Sources: 12

Artificial intelligence startup Anthropic has hired Amir Salek, a foundational leader of Google's Tensor Processing Unit project, to spearhead its expansion into custom semiconductor development. Confirming the appointment, Anthropic stated that Salek will join its internal compute team and report directly to Head of Compute James Bradbury. Salek previously spent eight years founding and leading Nvidia's system-on-a-chip business before directing custom silicon infrastructure for Google Cloud. The strategic hire underscores the company's ambition to mitigate supply chain bottlenecks, reduce external hardware dependency, and optimize silicon architecture specifically for its Claude models. Industry reports indicate Anthropic has engaged Samsung Electronics regarding advanced 2-nanometer manufacturing processes. The initiative mirrors rival OpenAI's custom hardware push alongside Broadcom. While pursuing proprietary silicon, Anthropic continues maintaining large-scale procurement agreements with major cloud providers as it prepares for an initial public offering.

By Sarah Whitman | JQJO News

Timeline of Events

  • On January 10, 2013, Amir Salek joined Google to launch custom silicon operations.
  • On March 15, 2022, Amir Salek transitioned to Cerberus Capital Management as managing director.
  • On August 1, 2026, Anthropic published job listings for an in-house chip team.
  • On August 21, 2026, Bloomberg reported Anthropic hired Amir Salek for hardware division.
  • On August 21, 2026, Anthropic officially confirmed the recruitment of Amir Salek publicly.
  • On August 22, 2026, reports detailed potential manufacturing talks with Samsung Electronics.
  • On August 22, 2026, industry analysts noted similarities to OpenAI's custom silicon strategy.
  • On August 22, 2026, Anthropic maintains heavy reliance on existing Nvidia and cloud hardware.
  • In future months, Anthropic will scale its internal semiconductor engineering team significantly.
  • In coming years, proprietary Anthropic silicon enters commercial data center deployment phases.

News Intelligence

  • US tech firms accelerate hardware independence amid severe infrastructure demand.
  • Proprietary silicon could permanently alter enterprise cloud and hardware costs.
  • US semiconductor designers, hardware engineers, investors, and major cloud providers.
  • Track primary corporate announcements and specialized semiconductor trade reporting outlets closely.
Media Bias
Articles Published:
12
Right Leaning:
0
Left Leaning:
0
Neutral:
12

Explain Framing

Left: Focuses on labor competition and corporate market concentration issues. Center: Emphasizes supply chain independence and technological scaling capabilities. Right: Highlights capital expenditure efficiency and private market competitive positioning.

Primary Source

Bloomberg reported Anthropic hired Amir Salek for custom silicon https://www.bloomberg.com/news/articles/2026-08-21/anthropic-hires-google-tpu-pioneer-to-build-custom-ai-chips

Media Bias
Articles Published:
12
Right Leaning:
0
Left Leaning:
0
Neutral:
12
Distribution:
Left 0%, Center 100%, Right 0%
Explain Framing

Left: Focuses on labor competition and corporate market concentration issues. Center: Emphasizes supply chain independence and technological scaling capabilities. Right: Highlights capital expenditure efficiency and private market competitive positioning.

Primary Source

Bloomberg reported Anthropic hired Amir Salek for custom silicon https://www.bloomberg.com/news/articles/2026-08-21/anthropic-hires-google-tpu-pioneer-to-build-custom-ai-chips

Coverage of Story:

From Left

No left-leaning sources found for this story.

From Right

No right-leaning sources found for this story.

Related News

Comments

JQJO App
Get JQJO App
Read news faster on our app
GET