OpenAI and Broadcom unveil LLM-optimized inference chip
OpenAI and Broadcom have collaborated to create Jalapeño, a new AI inference chip designed specifically for large language models. This accelerator promises significantly better performance per watt compared to current state-of-the-art technology, making advanced AI faster, more reliable, and more accessible.
OpenAI and Broadcom have unveiled Jalapeño, OpenAI's first intelligence processor, an accelerator designed from the ground up for LLM inference. This collaboration aims to create a multi-generation compute platform to make advanced AI faster, more reliable, and more accessible. The chip's architecture reduces data movement and optimizes compute, memory, and networking resources for near-theoretical peak performance.
Jalapeño was delivered to OpenAI's leadership within nine months, showcasing a rapid development cycle. OpenAI designed the chip with a deep understanding of LLM fundamentals, informed by its roadmap of models and product needs. Broadcom and Celestica assisted in industrializing the platform, handling chip implementation, system integration, and scalable production.
Early testing indicates that Jalapeño will deliver substantially better performance per watt than current leading technology. This efficiency is crucial for OpenAI's long-term strategy to expand its full-stack platform, from models to products and now to chips. The chip is designed for flexibility, compatible with all LLMs, and informed by OpenAI's insights into the inference needs of current and future AI models.
This initiative strengthens OpenAI's continuous improvement cycle. Enhanced infrastructure drives computational efficiency, leading to better training and serving of AI models. More capable models result in improved products for users, developers, and businesses, generating revenue that can be reinvested into further infrastructure development. This cycle ultimately makes intelligence more powerful, reliable, and affordable for everyone.
Related articles
OpenAI’s Jalapeño chip is Big Tech’s spiciest move away from Nvidia
OpenAI is challenging Nvidia's dominance in the AI chip market with its new custom inference chip, Jalapeño. This move positions OpenAI alongside other tech giants like Google and Apple, who are developing their own silicon to reduce reliance on single suppliers and gain more control over hardware performance.
AlgoEvolve: LLM-driven Meta-evolution of Algorithmic Trading Programs
AlgoEvolve introduces a novel approach to algorithmic trading by leveraging large language models (LLMs) for meta-evolution. This method allows for the creation of more adaptive and efficient trading programs. The research explores the potential of LLMs to revolutionize financial trading strategies.
Agentic Analysis for Agentic Infrastructure: An LLM-Powered Pipeline for Comparative Governance of DAO and Corporate AI Protocols
This research paper explores an LLM-powered pipeline for the comparative governance of Decentralized Autonomous Organizations (DAOs) and corporate AI protocols. It introduces an agentic analysis approach to understand and compare the regulatory frameworks of these distinct AI infrastructures.