Source: X
OpenAI is moving beyond software and inference into silicon design, partnering with established chipmakers MediaTek and Qualcomm rather than attempting in-house fabrication like Nvidia or Apple. The strategy embeds OpenAI's inference stack directly into devices, reducing latency and dependency on cloud endpoints for core AI features. Production hits mass market in 2028. Luxshare's involvement—Foxconn's manufacturing partner—signals intent to control the full supply chain. For OpenAI, hardware is now essential to competitive AI distribution, not peripheral.