OpenAI and Broadcom Unveil Jalapeño, OpenAI's First Custom AI Chip
OpenAI announced Jalapeño, its first custom artificial intelligence chip designed in partnership with Broadcom specifically for large language model inference workloads. The acceleration chip has reached the initial sample stage and is showing cost reductions of roughly 50% compared to standard AI graphics processing units, according to Broadcom CEO Hock Tan.
Developed from scratch to manufacturing tape-out in nine months, Jalapeño is optimized to run OpenAI’s ChatGPT, Codex, and API services. Broadcom and Celestica, which handles board and system integration, will prepare the silicon for deployment across data centers backed by Microsoft starting this year. OpenAI frames the release as the launch of a multi-generation roadmap targeting gigawatt-scale infrastructure, with a successor chip slated for 2028.
From the sources (25 posts)
@deitaone$AVGO - OPENAI AND BROADCOM UNVEIL LLM-OPTIMIZED INTELLIGENCE PROCESSOR
@wallstengineOPENAI AND BROADCOM $AVGO UNVEIL AI CHIP CALLED JALAPEÑO
@firstsquawkOPENAI, BROADCOM UNVEIL AI CHIP TO RUN MODELS FASTER, CHEAPER || OPENAI, BROADCOM UNVEIL LLM-OPTIMIZED INTELLIGENCE PROCESSOR
@firstsquawkOPENAI’S NEW CHIP SHOWS 50% COST SAVINGS, BROADCOM CEO SAYS || OPENAI & BROADCOM UNVEIL LLM-OPTIMIZED INTELLIGENCE PROCESSOR
@negligible_cap*OPENAI, BROADCOM UNVEIL AI CHIP TO RUN MODELS FASTER, CHEAPER $AVGO and OpenAI have reportedly developed a custom AI chip called Jalapeno, which shows cost savings of 50% compared to typical AI GPUs and will be integrated into $MSFT datac
@wallstengineOpenAI and Broadcom $AVGO unveiled Jalapeño, a custom AI accelerator built for LLM inference. The chip is expected to deploy with Microsoft and other data center partners starting in 2026, with Celestica handling board, rack, and system in
@negligible_cap"The companies said Wednesday that OpenAI has received the first samples of the chip, called Jalapeno, and is testing how the silicon handles running AI workloads. So far, the accelerator is showing cost savings of roughly 50% compared with
@stockmktnewzOPENAI AND BROADCOM $AVGO JUST UNVEILED OPENAI'S FIRST CUSTOM AI CHIP It is called Jalapeño. It is an inference accelerator built from scratch specifically for large language models. Designed from scratch to manufacturing tape-out in nine
@businessOpenAI unveiled its first custom artificial intelligence chip developed in partnership with Broadcom, part of a bid by the ChatGPT maker to gain an edge by tailoring hardware to better run its AI products.
@financialjuiceOpenAI & Broadcom AI Chip Announcement $AVGO OpenAI and Broadcom unveil an LLM-optimized intelligence processor OpenAI and Broadcom unveil an AI chip designed to run models faster and cheaper OpenAI's new chip shows 50% cost savings, accor
@stocksavvyshay$AVGO and OpenAI unveiled Jalapeño which is OpenAI’s first custom AI chip built for LLM inference. Broadcom says this is the start of a multi-generation roadmap for gigawatt-scale AI data centers with $MSFT and other partners beginning in
@cnbcOpenAI unveils first chip as part of Broadcom deal in effort to 'build the full stack'
@wallstengine$AVGO Hock Tan: OpenAI has received the first samples of the chip, called Jalapeno, and is testing how the silicon handles running AI workloads. So far, the accelerator is showing cost savings of roughly 50% compared with typical AI grap
@jukan05
@openaiWe’ve designed and built our first AI chip: Jalapeño. Designed from the ground up by OpenAI and brought to production with @Broadcom, Jalapeño is purpose-built for the LLM workloads powering ChatGPT, Codex, the API, and future agentic prod
@openai
@altryneJalapeño - open AI's first chip!
@jukan05I wonder how EDA companies will interpret this. Do you think OpenAI will try to sell these solutions externally?
@techmemeOpenAI and Broadcom unveil Jalapeño, an LLM-optimized inference chip developed from design to manufacturing tape-out in nine months, aided by OpenAI's models (OpenAI) (Visit Techmeme dot com for the link and full context!)
@andrewcurran_OpenAI built their own chip - the Jalapeño - designed for inference. They did it in just nine months. Quoting from the blog: 'OpenAI designed the chip from scratch around its deep understanding of LLM fundamentals, informed by its roadma
@kimmonismusOpenAI just unveiled Jalapeño, its first custom AI chip designed from scratch for LLM inference- It is OpenAI moving deeper into the full stack: chips, kernels, memory, networking, racks, scheduling, deployment and product experience. Ope
@reach_vbholy shitt!
@scaling01RT @OpenAI: We’ve designed and built our first AI chip: Jalapeño. Designed from the ground up by OpenAI and brought to production with @Br…
@thetranscript_Broadcom and OpenAI unveiled "Jalapeño," OpenAI’s first custom LLM inference accelerator, designed specifically for AI inference workloads and expected to deliver materially better performance-per-watt than current state-of-the-art systems.
@axiosOpenAI tests homegrown AI chips