Why everyone from OpenAI to SpaceX is building their own chips (and turning up the heat on Nvidia)

OpenAI announced plans to develop its own custom inference chip, codenamed Jalapeño, in collaboration with Broadcom, on March 14, 2024. This move signifies a broader industry trend where major technology companies are designing proprietary chips to reduce reliance on single suppliers like Nvidia, which has historically dominated the AI chip market. Google, Apple, and SpaceX are among the other prominent organizations actively developing their own custom silicon for AI workloads. The primary objective for these companies is to gain greater control over their hardware, optimize performance for specific AI tasks, and potentially lower costs associated with high-demand AI processing units. This strategic shift is intensifying competition and challenging Nvidia's long-standing market leadership in the AI semiconductor sector. The development of custom chips allows companies to tailor hardware to their unique software needs, potentially leading to more efficient and powerful AI systems. This diversification in chip manufacturing is expected to reshape the AI hardware landscape in the coming years.