OpenAI Chip Core Engineer Chan Leaves to Join Anthropic
0xBroomberg
Clive Chan, the second employee on OpenAI's hardware team, joined Anthropic this week — just as the custom AI accelerator he led into production with Broadcom hits the manufacturing stage. The timing signals Anthropic's own chip effort is shifting from idea to execution.
Why does this person matter?
Chan joined OpenAI in January 2024 as the second hardware employee and earliest independent technical contributor on its custom-chip project.
His core work: a custom AI accelerator (ASIC — a chip designed for one specific task, unlike a general-purpose GPU) built with Broadcom on TSMC's 3 nm process.
This means → he is not a rank-and-file engineer but a critical-path figure with end-to-end experience from design through mass production.
Why leave at exactly this moment?
Last October Chan said the chip would enter mass production in roughly nine months — a timeline that lines up almost exactly with his departure.
OpenAI's roughly 40-person chip team continues under its first employee, former Google TPU lead Richard Ho.
In plain terms = the hardest "zero-to-one" phase is done. Chan chose to leave during the wrap-up and start a new zero-to-one at Anthropic.
What does Anthropic gain?
In April, Reuters reported Anthropic wanted to develop its own chips, but the plan was still early-stage — no dedicated team, no confirmed design.
Before OpenAI, Chan worked on Tesla's Dojo team, building the software framework and data-center co-design for training ASICs, reporting directly to Musk each week.
This means → Anthropic picks up OpenAI chip-to-production plus Tesla Dojo experience in a single hire. Its chip plan will most likely shift from exploration to real execution.
What does this mean for the rivalry between the two companies?
OpenAI's valuation has reached the trillion-dollar range; Anthropic recently rose to $965 billion. Both are preparing IPOs.
In his departure post, Chan cited three reasons for choosing Anthropic: talent, values, ambition — and compared the move to "starting a climb from the base of a new peak."
This reflects a deeper signal: in the AI arms race, in-house chip capability is becoming the key differentiator among top companies — and where the talent flows is the direction marker.
Content is for reference only, not financial advice.