Broadcom Inc. (AVGO) and OpenAI announced the launch of Jalapeño, OpenAI’s first custom AI inference processor, specifically designed for large language models (LLMs). The chip, developed in nine months, is optimized for AI inference and has shown early tests indicating industry-leading performance per watt. The companies plan to deploy Jalapeño in gigawatt-scale data centers starting in late 2026, with Broadcom committed to supplying 1.3 gigawatts of compute in 2027 as part of a broader 10-gigawatt agreement through 2029.
Broadcom’s collaboration with OpenAI aims to reduce AI compute costs and enhance efficiency across current and future LLMs. The company’s management released figures from their second-quarter fiscal 2026 earnings call, revealing that silicon for OpenAI has been delivered and remains on track for production. Additionally, Broadcom anticipates AI semiconductor revenues will exceed $100 billion by 2027, fueled by demand from OpenAI and other hyperscale partners.
Despite these advancements, Broadcom faces stiff competition from NVIDIA and AMD in the semiconductor market. NVIDIA reported a surge in data center revenues, increasing to $75.2 billion in Q1 fiscal 2027, while AMD’s EPYC processors are expected to drive a revenue growth of over 70% year over year in Q2 fiscal 2026.
5 Stocks Our Experts Predict Could Double In the Next Year
By submitting your email, you'll also get a free pivot & flow membership. A free daily market overview. You can unsubscribe at any time.
