OpenAI Hired Microsoft Hardware Executive Wilson Vu

The appointment marks a shift as OpenAI moves toward internal silicon development for its compute architecture.

Updated on Oct. 12, 2026 in Semiconductors

OpenAI Hired Microsoft Hardware Executive Wilson Vu

Live Poll

Does the movement of senior AI talent between major companies increase your trust in AI development?

Wilson Vu has joined OpenAI as a member of the technical staff focusing on compute architecture. He departs Microsoft after a 12-year tenure where he served as senior director of AI systems hardware architecture.

Why it matters

Vu's transition brings specialized experience in large-scale hardware systems to OpenAI as the firm attempts to move beyond its current reliance on third-party silicon from Nvidia, AMD, and Google.

OpenAI is developing in-house AI hardware known as Jalapeño chips. The current compute fleet relies on external GPUs and TPUs, marking a transition toward custom, vertically integrated silicon.

The players

Wilson Vu

Former Microsoft senior director of AI systems hardware architecture who joins OpenAI's technical staff.

OpenAI

An AI research and deployment organization currently scaling its infrastructure beyond third-party hardware providers.

Microsoft

A cloud and software enterprise that maintains a significant compute infrastructure and partners with OpenAI.

The details

Vu joins OpenAI to support the development of Jalapeño, the company's internal AI hardware project. This effort requires managing complex compute architecture, which integrates processors designed for the specific memory and bandwidth demands of large language model training.

2026: OpenAI expects to deploy Jalapeño chips in limited quantities.
2027: OpenAI expects to deploy Jalapeño chips more widely.

The Tech Race

OpenAI's pivot toward internal hardware mirrors the trajectory of Google's TPU program. The company aims to exert more control over its compute stack as it prepares for the limited rollout of its first in-house silicon in 2026.

The development of Jalapeño chips will not immediately change service availability for users. The hardware is currently in development with limited deployment targeted for 2026 and broader scaling scheduled for 2027.

The takeaway

OpenAI is aggressively building an internal hardware capability to secure its long-term compute capacity. Stakeholders should monitor the 2026 limited deployment window for initial performance data regarding the Jalapeño chip architecture.