Nvidia and Amazon Fund Massive Texas Power Projects

Nvidia is reportedly planning to invest up to $3 billion in Lancium, a power infrastructure developer that already supports data centers for OpenAI and Oracle in Texas. A $2 billion portion of this investment would secure Nvidia an approximate 20 percent stake in Lancium, valuing the infrastructure firm at about $10 billion. Lancium currently has four gigawatts of power under contract in Texas and is actively developing additional sites to secure up to 15 more gigawatts of capacity.

Meanwhile, Amazon is backing a major gas-fired power plant project in Pecos County, Texas, to fuel its own infrastructure. The planned facility will feature 35 gas turbines designed to generate up to 7.65 gigawatts of electricity dedicated to an Amazon data center. However, this fossil-fuel approach carries a heavy environmental toll, with projections indicating the plant could emit up to 33 million tons of carbon dioxide annually, potentially making it the highest-emitting power plant in the United States. Amazon spokesperson Margaret Callahan acknowledged that while the company's climate goals remain in place, the energy demands of artificial intelligence data centers make achieving those targets more difficult.

For AI practitioners and developers, these massive infrastructure plays highlight a shifting bottleneck in the industry. As models grow larger and agentic AI becomes more common, the constraint is no longer just about acquiring GPUs, but securing the raw electrical power to run them. Climate scientist Zeke Hausfather estimates that a single heavy user of agentic AI consumes roughly as much energy per year as a household clothes dryer, a figure that continues to rise. Developers must increasingly weigh the computational capabilities of next-generation models against the mounting environmental costs and regional grid constraints associated with training and deploying them.