OpenAI and Broadcom unveiled their first custom AI chip, called Jalapeño, marking OpenAI’s entry into the artificial intelligence silicon space on Wednesday.
The new chips will be manufactured by Broadcom and used by OpenAI for inference—the compute-intensive process of serving its AI models to users across ChatGPT and other applications.
In a launch press release, OpenAI President Greg Brockman said, “By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.” OpenAI has highlighted that it developed the Jalapeño chip from the ground up, drawing on its deep understanding of LLM fundamentals.
Throughout the development process, the company was guided by its roadmap for models, kernels, serving systems, and product requirements.
Partners Broadcom and Celestica helped industrialize the platform through chip implementation, board and rack integration, high-performance networking, and scalable production systems.
Engineering samples of the chip are already running machine-learning workloads in the lab at target production performance and power levels, including GPT-5.3 and Codex Spark.
More from Tech Rockstar reveals GTA VI pricing, offers Vice City-themed bonuses for early purchases Nothing phone (4b) specs surface online ahead of launch: Here's what to expect Designed to Be the Best Inference Platform OpenAI describes Jalapeño as a blank-slate design built specifically for modern LLM inference, rather than a general-purpose accelerator adapted from earlier AI workloads.
The chip was developed using insights gained from ChatGPT, Codex, the OpenAI API, and future agentic products, while also being optimized for current and next-generation LLMs across the industry.
The goal is to combine the throughput of today’s leading AI accelerators with latency closer to that of specialized inference systems, making Jalapeño well-suited for interactive LLM deployments at scale.
Quick Reads View All Samsung Galaxy M47 set for India launch on June 29: Display, battery, processor and more Chinese EV makers are shut out of India - but their tech isn't Jalapeño strengthens OpenAI’s full-stack AI strategy by improving compute efficiency, enabling the company to build faster, more capable, and more affordable AI models while reinvesting those gains into future infrastructure.
