$1.3 trillion evaporated from the chipmaker market: the real problem is not a chip shortage, but electricity and transformers
Over the past month, shares of leading semiconductor manufacturers have lost more than $1.3 trillion in market capitalization. However, the root cause of this crash is not an oversupply of chips or a drop in demand for them. As my analysis shows, the market is misinterpreting the signals — the bottleneck for AI infrastructure development is not "silicon," but "iron" in the most literal sense: steel, copper, and energy.
Amid a massive sell-off, Taiwan Semiconductor Manufacturing Co (TSMC) reported a record quarter. The company's revenue reached $40.2 billion, its growth forecast for the next period was raised by more than 40%, and management described confidence in the artificial intelligence megatrend as "very high." TSMC's profit surged by 77%, hitting an all-time high, and its margin exceeded all previous figures. The company also announced an additional $100 billion in investments for its plants in Arizona.
As I see from the data, the paradox is that the sector's crash and the stellar report cannot both be correct regarding demand. Reality reflects the report: chips are being shipped, demand is accelerating, not fading. The problem has shifted from the realm of semiconductor production itself to a completely different area.
Where is the real "bottleneck"?
The chips themselves are no longer a scarce resource for AI. Now, the industry's development is constrained by steel, copper, and time. The first bottleneck is chip packaging. The capacity of CoWoS technology, which connects the processor to memory, is already insufficient: Nvidia buys up about 60% of these lines, and demand for packaging has tripled in two years. TSMC is nearly doubling its production in this area, but the lines are still operating at full capacity.
The second and much more complex constraint is energy. A finished chip is useless without electricity, and the industry is acutely short of it. Data center statistics confirm this energy hunger. Companies have announced 16 gigawatts of capacity in the US for 2026, but are only building 5 of them: the remaining projects are frozen, and 25% do not even have an energy supply plan.
The root of the problem, in my deep conviction, lies in equipment that the market hardly mentions. This refers to high-voltage transformers that convert grid energy into data center power. Currently, their manufacturing takes an average of 48–60 months, whereas before 2020, 12 months sufficed.
The shortage will persist for a long time. Hitachi, Siemens Energy, GE Vernova, and ABB — the four manufacturers of such installations — have accepted orders years in advance. At Siemens Energy alone, the order backlog has reached nearly €136 billion. The problem starts at the raw material level: special electrical steel for transformers is produced by only five companies in the world, and they cannot quickly ramp up production.
Circumventing the shortage is also not possible. Some data center operators decided to supply their own energy and ordered gas turbines, but GE Vernova, Siemens, and Mitsubishi have already sold their capacity for about five years ahead and more.
Analyst's conclusions: the AI story is now about electricity
To summarize the timelines for each constraint: chips are produced without delays, packaging capacity is fully booked, new transformers will appear in 2029, and turbines in 2030. The largest construction project in economic history is being held back by heavy electrical equipment that is produced almost manually.
This picture changes the perspective on recent sell-offs. Traders were dumping shares out of fear of falling demand, but the data points to a supply shortage. Beneficiaries could include transformer and turbine manufacturers, grid companies, and owners of sites with ready energy connections.
My professional assessment: the world's most advanced companies are forced to build their own power plants to run neural networks. The AI story is now not about intelligence, but about electricity: the market was counting chips, when it should have been counting substations. For long-term growth-oriented investors, it is worth shifting focus from semiconductor giants to energy infrastructure companies — they will become the main beneficiaries of the next wave of AI development.