OpenAI has unveiled its first custom-designed AI accelerator, Jalapeño, built to deliver faster and more efficient LLM inference. Developed in collaboration with Broadcom, the chip is described as an “Intelligence Processor” created specifically for inference workloads rather than AI model training.
The company said Jalapeño will power services including ChatGPT, Codex, the OpenAI API, and future agentic AI products. Unlike conventional AI accelerators adapted from general-purpose designs, Jalapeño has been purpose-built for AI inference. Its architecture, board layouts, rack integration, networking, and deployment systems have all been developed jointly with Broadcom, while Celestica is the manufacturing partner.
According to OpenAI, engineering samples are already running machine learning workloads at production target frequencies and power levels in its laboratories, including GPT-5.3-Codex-Spark.
The company claims Jalapeño improves performance by reducing data movement and balancing compute, memory, and networking resources, enabling hardware utilisation much closer to its theoretical limits. However, OpenAI has not yet released benchmark performance figures.
OpenAI President and Co-Founder Greg Brockman said the chip is part of the company’s broader strategy to make AI faster, more reliable, and more affordable, adding that the world is rapidly moving towards a “compute-powered economy.”
OpenAI expects Jalapeño to become the foundation of a multi-generation AI computing platform starting next year.
Broadcom and OpenAI also revealed that Jalapeño progressed from the initial design stage to manufacturing tape-out in just 9 months, making it one of the fastest ASIC development cycles for a high-performance AI semiconductor. Broadcom contributed its silicon implementation expertise and high-speed networking technologies, including its Tomahawk networking silicon, while OpenAI will deploy the platform across large-scale AI data centres.
The launch of Jalapeño highlights OpenAI’s long-term strategy to build its own AI infrastructure instead of relying entirely on third-party AI hardware.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter
About us:
The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.