OpenAI and Broadcom unveil Jalapeño inference chip

OpenAI and Broadcom unveiled Jalapeño, OpenAI’s first Intelligence Processor and the first accelerator in a multi-generation compute platform aimed at LLM inference. The chip was designed around OpenAI’s model roadmap, serving systems, kernels, and product needs, with Broadcom contributing silicon implementation and networking technologies and Celestica supporting board, rack, and system integration.

Engineering samples are running ML workloads in the lab at production target frequency and power, including GPT‑5.3‑Codex‑Spark. OpenAI said final measurements are still underway, but early testing shows performance per watt substantially better than current state-of-the-art. The architecture is intended to reduce data movement and balance compute, memory, and networking resources so workloads run closer to theoretical peak performance.

Jalapeño was co-developed from initial design to manufacturing tape-out in just nine months, with OpenAI models used to accelerate parts of the design and optimization process. The platform is designed for initial deployment by the end of 2026 and is expected to expand over multiple generations, including gigawatt scale data centers with Microsoft and other partners.