Broadcom Unveils 'Jalapeño' AI Intelligence Processor in OpenAI Partnership
Key Facts
In a move that solidifies its position in the global AI hardware race, Broadcom has officially unveiled 'Jalapeño', its new custom-designed intelligence processor. This announcement comes eight months after the initial disclosure of a custom silicon deal between Broadcom and OpenAI. The partnership is strategically aimed at providing OpenAI with specialized hardware infrastructure to power its next-generation AI models while optimizing operational efficiency.
The unveiling of 'Jalapeño' places Broadcom in direct competition with semiconductor giants like Nvidia and Marvell Technology, who currently lead the custom ASIC market. According to market data, hyperscalers and AI labs are increasingly pivoting toward bespoke silicon to mitigate high costs and supply constraints, a trend that Broadcom is uniquely positioned to capture. This hardware milestone underscores the company's transition from a networking specialist to a core AI infrastructure provider.
Regarding market performance, AVGO stock stood at $380.15 (close June 23, 2026), having reached an intraday high of $391.02 per market data. Traders are currently monitoring support levels near $376.66 to gauge price stability following the product news. Looking ahead at the economic calendar, while no direct corporate catalysts are scheduled for the next seven days, technology sector sentiment remains sensitive to the Federal Reserve's interest rate path confirmed on June 17.
Latest Updates · 2
- Notable·
Update: This partnership specifically targets lowering ChatGPT's operational costs through optimized LLM inference, achieving a rapid nine-month design-to-tape-out cycle. Driven by superior performance-per-watt metrics, OpenAI internal projections now suggest a path to profitability between 2029 and 2030.
- Notable·
Update: Celestica has emerged as a key infrastructure partner in the project, providing the specialized rack technology required for the new processors. Furthermore, the technical scope of the Jalapeño chip has been clarified as an AI inference processor, specifically optimized for executing trained models with high efficiency for OpenAI.