OpenAI reportedly planning to build its first AI chip in 2026
SMRTR summary
OpenAI is developing its own AI chip with help from TSMC and Broadcom, aiming for a 2026 release. The company has shifted focus from building chip factories to in-house design. To address chip shortages, high costs, and delays, OpenAI is also diversifying its training hardware by using AMD chips through Microsoft's Azure cloud platform, in addition to Nvidia GPUs. These moves could help OpenAI reduce dependence on a single supplier and potentially lower AI training costs.
SMRTR provides this summary for quick context. The original article belongs to TechCrunch.
Read the original article