OpenAI Denies Report of Dissatisfaction With Nvidia AI Chips
OpenAI Denies Report of Dissatisfaction With Nvidia AI Chips
Publish Date: 2026-02-03 12:51:00
Source Domain: www.pymnts.com
Here is a respectful and informative summary of the article in an unordered list format, highlighting between 4 and 8 key points:
– Sam Altman, CEO of OpenAI, publicly expressed strong satisfaction with Nvidia’s AI chips, stating that Nvidia makes the best AI chips and that they aim to remain a significant customer for a long time.
– Contradicting earlier reports by Reuters, Altman dismissed claims of dissatisfaction, questioning the sources of the “insanity” surrounding these reports.
– According to Reuters, sources revealed that OpenAI had been unsatisfied with the performance of Nvidia’s chips, specifically regarding speed, hence exploring alternatives from companies like AMD, Cerebras, and Groq.
– Despite the reported dissatisfaction, an official statement from an OpenAI spokesperson confirmed that Nvidia powers most of their inference fleet and offers the best performance per dollar.
– As per Reuters’ report, it’s mentioned that OpenAI announced plans to integrate 750 megawatts of ultra-low latency compute from Cerebras to enhance response times for AI models.
– It was revealed that a partnership between AMD and OpenAI involves developing AI data centers utilizing AMD processors, indicating a shift or diversification in chip suppliers.
– The article touched upon controversy surrounding Nvidia’s proposed $100 billion investment in OpenAI, with internal doubts and subsequent clarifications from Nvidia CEO Jensen Huang asserting ongoing support and investment.
These points provide a clear overview of the key narratives discussed in the article surrounding the relationship between OpenAI, Nvidia, and other chip suppliers.
