OpenAI Explores Chip Alternatives Amid $100B Nvidia Deal Delays
OpenAI is actively diversifying its hardware suppliers, seeking alternatives to Nvidia's dominant AI chips for inference tasks. The ChatGPT Maker has struck deals with AMD, Broadcom, and Cerebras Systems to address performance limitations in about 10% of its inference workloads.
The stalled $100 billion investment deal with Nvidia reflects growing tensions in the AI hardware space. While both CEOs publicly maintain cordial relations—with Sam Altman calling Nvidia's chips 'the best in the world'—the competition is intensifying. Nvidia has responded by licensing Groq's technology for $20 billion and poaching its engineers.
At stake is the future of AI inference, where response speed dictates user experience. OpenAI's search for chips with embedded memory highlights the industry's shift toward specialized hardware as large language models evolve.
