Intel has refocused its AI efforts on inference, conceding Nvidia’s dominance in AI training. The company’s Gaudi line failed due to immature software and unfamiliar architecture. CEO Lip-Bu Tan admitted catching up to Nvidia is unlikely. Intel aims to excel in AI inference and agentic AI, targeting edge data centers and devices. The company may abandon rack-scale AI solutions, focusing on smaller, cheaper AI models. A turnaround for Intel will be a long process.

Read more at Nasdaq: Intel Might Be Quitting the AI Training Market for Good