TSMC worried about chip oversupply? Musk: They are right, power and liquid cooling can't keep up

Recently, Musk stated that TSMC's concerns about chip overcapacity are "correct." He predicts that the limiting factor in the AI industry will shift from chip manufacturing to "getting the chips up and running," with the core bottleneck being the supply of electricity, transformer configuration, and the deployment of cooling systems. On January 6, Musk engaged in an in-depth dialogue about the future of AI with Peter Diamandis, Executive Chairman of Singularity University, and David Blundin, founder of Link Ventures. He pointed out that current chip production capacity is growing exponentially, while the energy infrastructure supporting its operation can only expand linearly. When the two curves intersect, a large number of high-performance AI chips will be unable to be utilized due to a lack of supporting power conversion equipment and cooling systems. This judgment directly addresses the issue of the current AI infrastructure construction being severely underestimated. For investors, this means that the focus of the AI computing power race is shifting from chip procurement to the capability of building energy infrastructure. Electricity Becomes the "Rate Limiter" for AI Deployment During the dialogue, Musk further revealed the specific forms of the electricity bottleneck in AI infrastructure. He emphasized that deploying AI chips is far from as simple as "bringing GPUs to the power plant," but requires simultaneously addressing three core issues: gigawatt-level power supply, high-voltage power conversion, and efficient cooling systems. Musk specifically pointed out that the entire data center industry is undergoing a critical transformation from air cooling to liquid cooling, and warned that this process carries significant risks. He stated: “This is a fundamental shift for data centers, which have long relied on air cooling. If a liquid cooling pipe bursts—like a water pipe bursting in a data center—it could lead to a loss of $1 billion.” He used xAI's "Giant 2" project in Memphis as an example to illustrate the actual challenges: although the project site is located near multiple 300 kV high-voltage lines, completing the connection still takes about a year. To achieve the launch of a 1-gigawatt training cluster by mid-January 2025, the team had to temporarily combine multiple gas turbines ranging from 10 to 50 megawatts as a transitional power source and use a large number of Megapack battery packs for power smoothing. The "Curve Intersection" of Chip Production Capacity and Power Supply When asked whether he agrees with TSMC's concerns about overcapacity, Musk gave an affirmative response: “I’m not sure about their reasoning, but the conclusion is correct.” He pointed out that the key is to identify the "limiting factors of each period" and predicted that by the third quarter of 2026 (about 9-12 months from now), the core bottleneck will shift from chip manufacturing to the ability to "get the chips running." This judgment stems from the misalignment of two development trajectories: AI chip production capacity is expanding at an exponential rate, while the supporting power infrastructure can only grow linearly. Musk emphasized: “If chip output is growing exponentially while power supply can only slowly increase linearly, the two curves will inevitably intersect.” This means that the speed of chip manufacturing may far exceed the speed at which they can actually be deployed and powered on. In this regard, David Brendin expressed a different view, believing that even if Taiwan Semiconductor increases GPU production from 20 million to 40 million units, the market will still find a way to address the power supply issue. However, Musk insisted that any lack of power conversion links or cooling systems would prevent the chips from being truly activated, fundamentally suppressing actual demand and purchasing behavior