The former chief executive of Intel has delivered a blunt verdict on the semiconductor industry: the vast majority of chip design firms will vanish, according to comments reported by 36Kr. In his judgment, manufacturing stands as the only core moat.

The statement carries a certain weight of biography. Pat Gelsinger spent his tenure at Intel betting tens of billions of dollars on fab construction, an IDM strategy that priced fabrication as the irreducible core of the industry. His latest remarks, as summarized by 36Kr, extend that logic to its conclusion. Design, in this view, is a crowded and ultimately commoditized discipline. Manufacturing is not.

The claim is a forecast, not a measurement, and readers should treat it as one man's position rather than an industry consensus. No survey, shipment figure, or attrition rate accompanies the judgment in the report. What it does offer is a framing of the competitive structure of the chip business, and that framing deserves scrutiny on its own terms.

Why design faces the squeeze?

The argument rests on a simple asymmetry. Design firms multiply easily; advanced fabs do not. A design startup needs engineers and a foundry partner with an available process slot. A leading-edge fab needs capital in the tens of billions, years of construction, and a customer base deep enough to fill the tool.

The consequence, in the former CEO's framing, is that design differentiation erodes while manufacturing scarcity persists. Firms without a defensible position will consolidate or disappear. The survivors will be those anchored to the physical layer of the supply chain, where the barriers remain prohibitive.

What does this mean for the foundry race?

Taken at face value, the judgment favors the small set of companies that actually run leading-edge fabs: Intel itself, TSMC, and Samsung. It also reads as a posthumous defense of the Intel Foundry strategy that Gelsinger championed and that his successors have had to defend to investors. A moat argument made by the man who dug the moat is at least convenient.

For the design ecosystem, the implication is consolidation. For the manufacturing layer, it implies continued pricing power and strategic leverage, which is precisely why governments in the United States, Europe, and Asia have subsidized fab construction so heavily in recent years.

How much should one prediction move anyone?

A single executive's forecast, however experienced, is not data. The report offers no timeline for the predicted disappearance of design firms, no threshold for what counts as "vast majority," and no count of the firms at risk. Analysts have argued both sides of the design-versus-manufacturing question for a decade, and the industry currently sustains thousands of fabless companies, a fact that sits awkwardly beside any imminent-extinction thesis.

Still, the remark lands at a moment when advanced packaging and leading-edge capacity are genuinely scarce, and when design tooling costs at the newest nodes favor only the largest players. The economics that make design harder are real, even if the mortality forecast is arguable.

The market will render its own verdict over the coming years, through foundry utilization rates, consolidation counts, and the pricing of advanced-node wafer capacity. Those numbers, unlike predictions, keep no counsel.