Democrats' US midterms victory is a win for hyperscalers and AI labs, a mixed bag for Nvidia, an advantage to China, and may make space data centers attractive

Data center construction permits may become harder to secure across the US.

Democrats secured a significant victory in the US midterm elections, gaining control of both chambers of Congress. This outcome was influenced by voter discontent over the economy and the war in Iran. A notable trend during the election cycle was the prominent discussion of data centers in political campaigns, with both parties expressing opposition. In Texas, a key battleground state, the election of Gina Hinojosa as governor signals potential restrictions on AI data centers. Texas accounts for over 25% of planned US data center capacity by 2030, making this development significant for the tech industry. This political shift suggests that the expansion of AI infrastructure may face substantial political hurdles, potentially becoming the most significant constraint on its growth.

The election results and the strong anti-data center sentiment, particularly in states like Texas, indicate a substantial political roadblock for AI infrastructure expansion. This opposition, fueled by concerns over job displacement and environmental impact, could lead to prolonged scarcity of compute power and memory. Hyperscalers are poised to benefit as they possess existing sites and greater capacity to navigate regulatory landscapes, potentially increasing their market share in AI compute. Conversely, 'neoclouds' may struggle due to the difficulty in obtaining new permits and capital.

Nvidia faces a mixed outlook. While a slowdown in data center construction could reduce demand for its chips, the overall constraint on AI infrastructure might make its high-efficiency offerings more attractive. The company is actively positioning its products as essential for maximizing compute within limited power and data center footprints. However, hyperscalers developing their own AI chips pose a long-term competitive threat, especially if they can achieve comparable or superior efficiency and cost-effectiveness.