Amazon and Nvidia report strong AI data center demand despite recession fears

In Oklahoma City, Amazon and Nvidia executives addressed concerns regarding the artificial intelligence sector and its impact on data center construction. During a recent conference organized by the Hamm Institute for American Energy, executives from both companies expressed confidence in the demand for AI data centers, which has not waned despite recession fears. Kevin Miller, Amazon's vice president of global data centers, noted that there has been no significant decline in demand, with expectations that this trend will continue to grow both in the short and long term. His statements are noteworthy when juxtaposed with growing Wall Street anxieties over possible pullbacks in data center development from major tech companies. Recent reports indicated that Amazon Web Services is reportedly pausing some leases on data center commitments, drawing parallels to Microsoft’s recent reduction in early-stage projects. This led to fears of a broader pattern within the industry, potentially impacting future investments in AI technologies. Josh Parker, senior director of corporate sustainability at Nvidia, echoed Miller's sentiments, stating that they have not observed any signs of a slowdown in demand for data centers. The executives addressed the recent market reaction triggered by China's AI startup DeepSeek, which sparked sell-off in energy stocks. Investors expressed apprehension that DeepSeek’s advanced model might require less energy from data centers than previously anticipated. Parker dismissed this reaction as premature, arguing that the requirements for computing power and energy associated with AI are only going to rise. In light of these developments, Jack Clark, co-founder of Anthropic, emphasized the necessity for increased power capacity to accommodate the escalating needs associated with AI technologies, predicting that around 50 gigawatts of new power will be essential by 2027. This demand is comparable to the output of approximately 50 new nuclear plants. The conference served as a platform for discussing the urgent need for a collaborative effort between the technology and energy sectors to tackle the growing energy demands driven by artificial intelligence applications. There seems to be a consensus that natural gas will be a crucial component in meeting these new power requirements, as companies seek to bolster their infrastructure to support ongoing and future AI endeavors.