OpenAI has detailed its Jalapeño custom AI chip, disclosing a 700W thermal design power figure that places it among the most power-intensive processors yet announced for AI workloads. The specification signals a significant infrastructure challenge for any facility deploying the chip at scale, requiring purpose-built cooling and power distribution systems. Data Center Dynamics reported the details, which follow earlier reports that Jalapeño outperforms Nvidia Blackwell on certain inference tasks.
A 700W TDP per chip forces data center operators to redesign rack power budgets, cooling loops, and facility electrical infrastructure well beyond current norms. At scale, facilities housing thousands of Jalapeño units will face energy and thermal loads that challenge existing building standards and utility agreements.
Selected on the specific 700W TDP figure, the named chip 'Jalapeño,' and OpenAI as principal actor. The previously published story covered performance claims; this article adds the specific thermal specification, which is a distinct and consequential new detail. Ranked above other compute stories because hardware thermal specs directly drive infrastructure investment decisions.