The Consolidation Dynamics
Consolidation in the integrator tier is not an accident of ambition. It is the predictable output of the economics, and the economics point in one direction. Four forces push relentlessly toward a smaller number of larger players, and each one reinforces the others.
The first force is the learning curve. A factory that has built two hundred identical e-houses builds the two hundred and first faster, cheaper, and with fewer defects than a rival building its tenth. Cumulative volume drives down unit cost and drives up reliability. In a repetitive manufacturing process this advantage compounds, because the low-cost, high-reliability producer wins the next order, which adds to cumulative volume, which widens the cost and quality gap again. This is the classic experience-curve flywheel, and it structurally favors the largest incumbent. A fragmented field trade never accumulates this advantage because every project is a first build. A factory accumulates it with every unit.
The second force is component allocation, and in the current market it is decisive. The modules depend on inputs that are themselves scarce and slow to expand: large power transformers, medium-voltage switchgear, high-capacity chillers. When these components are rationed, the integrator that has secured allocation — through scale, long relationships, blanket orders, or outright ownership of the component maker — can promise delivery dates a smaller rival cannot meet at any price. Scale buys allocation, and allocation wins orders, which justifies more scale. Several integrators have moved to buy their component suppliers precisely to lock this up. Vertical integration converts a market dependency into a captive supply and denies it to competitors at the same stroke.
The third force is capital intensity. A modern module factory is an expensive, long-lead asset — eighteen to thirty-six months to design, build, staff, and qualify. Only a well-capitalized firm, or one backed by patient infrastructure capital, can build capacity ahead of demand. This raises the barrier to entry sharply. A new entrant cannot simply appear next quarter to relieve a shortage; the response time of supply is measured in years, which is exactly the condition under which incumbents keep pricing power for a long time. High capital intensity plus long build times means the supply curve is steep and slow, and steep-and-slow supply is the soil in which chokepoints grow.
The fourth force is the customer itself. Hyperscalers want fewer suppliers, not more — partners who can deliver at enormous scale, to a standard design, across many sites, with a single accountable throat to choke. A buyer standing up gigawatts of capacity cannot manage a long tail of boutique fabricators; it wants two or three integrators who can each deliver hundreds of modules to a repeatable specification. This preference actively selects for large integrators and starves small ones of the volume they would need to climb the learning curve. The demand side is, in effect, financing the consolidation of the supply side.
Put the four forces together and the trajectory is clear. Scale lowers cost and wins volume; volume buys component allocation and denies it to rivals; capital intensity keeps new entrants out for years; and the largest customers deliberately concentrate their spending on the biggest players. Each force feeds the next. The market moves from many interchangeable fabricators toward a handful of dominant integrators, and it does so faster than casual observers expect, because the flywheel accelerates as it turns.
The consolidation is visible in the deal flow. Switchgear and electrical majors are acquiring prefabrication houses to move downstream into finished modules. Cooling specialists are buying or being bought to add liquid-cooling capability as artificial-intelligence density makes it mandatory. Private capital is rolling up regional skid fabricators into national platforms. Component makers and integrators are merging in both directions. Every one of these transactions removes an independent link from the chain and hands more of the pipe to fewer hands. The through-line of all this activity is a single narrowing of the assembly-and-test layer.
See the pattern for what it is and the strategic implication writes itself. This is not a temporary shortage that new supply will soon relieve. It is a structural concentration driven by forces that reward size and punish fragmentation, unfolding in a layer that is slow and costly to widen. The integrator tier is consolidating into a chokepoint, and the consolidation is not a risk to be waited out. It is the base case to be planned around. The next lesson turns to what that means for the buyer standing on the wrong side of the narrowing pipe.