Latest AI News

OpenAI plans to produce its first custom AI chip with Broadcom by 2026, aiming to reduce dependence on Nvidia. This follows similar moves by Google, Meta, and Amazon. Broadcom has secured over $10 billion in orders for custom AI chips, with OpenAI as the reported client. The chip will be used internally to power OpenAI's AI models, not sold externally.

OpenAI is partnering with Broadcom and TSMC to mass produce custom AI chips by 2026 using 3nm technology. This strategic move aims to reduce dependency on third-party suppliers and optimize hardware for AI models like ChatGPT. Broadcom's AI revenue grew 63% year-over-year, with OpenAI contributing significantly to their $10 billion in new orders. This development reflects a broader industry trend where major tech companies are developing proprietary hardware to gain competitive advantages in the increasingly fragmented AI chip market.

OpenAI has partnered with Broadcom to mass-produce custom AI chips starting in 2026, reducing dependence on NVIDIA. With a $10 billion order commitment, this strategic alliance aims to optimize AI workloads and reshape the competitive landscape. Broadcom's Q3 2025 AI revenue reached $5.2 billion, while OpenAI projects $12.7 billion in 2025 revenue. This collaboration targets the growing AI hardware market, forecast to reach $360 billion by 2032.

Broadcom projects strong AI growth through fiscal 2026, reporting a 63% increase in AI revenue to $5.2 billion last quarter. A significant development is securing a fourth major customer for custom AI chips (XPUs) with $10 billion in new orders scheduled for 2026 delivery. The semiconductor solutions business saw 57% revenue growth, while infrastructure solutions jumped 43%. CEO Hock Tan will remain until at least 2030, providing leadership continuity as the company solidifies its position as a leader in custom silicon for hyperscale data centers, with Wall Street responding positively to these developments.