OpenAI and Broadcom have jointly unveiled “Jalapeno,” the first custom AI chip to emerge from their eight-month partnership – and it signals a major strategic shift in how the world’s most powerful AI company plans to power its future. The chip, which OpenAI is calling an “Intelligence Processor,” is designed specifically for inference: the compute-intensive task of serving AI model outputs to hundreds of millions of users across ChatGPT and other platforms. A physical sample was delivered to OpenAI on Wednesday, June 24, with initial deployment targeted for late 2026 and broader rollout planned into 2027.
The Jalapeno chip is the centerpiece of a broader platform OpenAI and Broadcom describe as being built “to make advanced AI faster, more reliable, and more accessible.” By designing custom silicon, OpenAI aims to reduce its dependence on third-party GPU providers like Nvidia and control more of its cost structure as its inference workloads scale exponentially. OpenAI President Greg Brockman framed it plainly: “By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.” Broadcom shares climbed roughly 2% on the announcement.
For investors, this is a significant signal. OpenAI joining the custom-silicon club – alongside Google (TPUs), Amazon (Trainium/Inferentia), and Microsoft (Maia) – validates that the hyperscale AI buildout is becoming a chip design race, not just a chip buying race. Broadcom (AVGO) is the key beneficiary as the manufacturing partner, and this deal adds a durable, long-term revenue stream to its AI custom chip portfolio. More broadly, it puts pressure on Nvidia’s dominance in inference workloads, which represent a growing share of total AI compute spend. Investors tracking the AI infrastructure theme should watch whether OpenAI’s chip ambitions accelerate or whether Nvidia retains pricing power – both stories are investable.
