Semiconductors: Moore’s Law and Rock’s Law

The semiconductor industry is a model of ongoing and rapid technological change, with an ongoing race between increased technological capabilities and higher costs of building production facilities. Brian Albrecht, Geoffrey A. Manne, David Teece, and Mario A. Zúñiga provide a readable overview in “From Moore’s Law to Market Rivalry: The Economic Forces That Shape the Semiconductor Manufacturing Industry “(International Center for Law & Economics, November 12, 2025).

Back in 1965, Gordon Moore, one of the founders of Intel, put forth the proposition that the number of transistors on a computer chip would double every two years. Moore’s law, as it became known, has turned out to be almost eerily accurate since then, which in turn has driven computing power that is vastly cheaper over time. (For earlier posts about Moore’s law around the time of its 50th anniversary, see here and here.) For semiconductor firms themselves, Moore’s law is also a threat: keep up or be left behind.

But at about the same time, Moore also enunciated a second and lesser-known law, which he attributed to an Intel board member named Arthur Rock. Rock’s law was that the cost of building a semiconductor fabrication plant would double every four years–that is, not nearly as fast as the technical capacities of the underlying semiconducter chips, but still quite a meaningful increase. A single advanced semiconductor fab now costs north of $10 billion. The authors note that for the chipmaker TSMC, capital expenditures are typically 30–50% of revenue, and if you include R&D, total firm investment often exceeds 40–60% of revenue. Moreover, every cutting-edge new computer chip is actually a bundle of new techologies: “a multi‑front physics problem involving (among other things) lithography, new transistor structures, novel materials, power delivery, metrology [methods of measurement], and advanced packaging.”

Thus, the competitive dynamics of the semiconductor industry are a race between expectations of ever-rising technological gains and ever-rising costs of building a plant. This is clearly not an industry for the faint of heart or the lightly capitalized.

The resulting industry structure has involved an evolution from chip firms that both designed and made chips to separate firms for chipo design and manufacturing: that is, some of the best-known “semiconductor” companies design chips, but do not actually make tthem. The authors write:

As the cost and technological demands of manufacturing at the bleeding edge became unsustainable for non-specialized firms, the development of standardized design tools and collaboration routines lowered coordination costs. These enabled a separation of design and manufacturing which, in turn, facilitated valuable specialization and risk-sharing: Independent foundries aggregate demand across multiple customers, achieving economies of scale difficult for IDMs [integrated design manufacturers] to reach, while also keeping their expensive fabrication equipment fully utilized.

This specialization created concrete economic benefits. Independent foundries achieve higher equipment utilization by serving diverse customers with different demand cycles. They spread enormous R&D costs across multiple clients rather than bearing them alone. The fabless model allowed companies like Nvidia to focus entirely on GPU architecture without operating fabs, while Qualcomm could specialize in wireless chips, and Broadcom in networking semiconductors. Meanwhile, each could access the same cutting-edge manufacturing technology that previously only giants like Intel could afford.

These sophisticated relationships are governed by relational contracts—contractual relationships whose precise terms depend on cooperative adjustments by the parties over time—and complex governance structures that solve extraordinary coordination challenges more efficiently than vertical integration or spot-market transactions. When capital investments reach tens of billions of dollars per facility and must be committed years before demand is certain, both foundries and their customers make relationship-specific investments that create bilateral dependence.

You can read the article by Albrecht, Manne, Teece, and Zúñiga for details on how these forces have shaped the evolution of the semiconductor industry. Here, I will just mention a few broader thoughts about For additional details on the chip industry, I’d also recommend the article by Chad P. Bown, and Dan Wang, “Semiconductors and Modern Industrial Policy” in the Fall 2024  Journal of Economic Perspectives (where I work as Managing Editor).

The “semiconductor industry” is not well-described as a single industry in which firms make reasonably similar products–unlike, say, the car industry. Instead, it’s an interwoven and overlapping mixture of equipment suppliers, chip designers, chip manufacturers, and end-users. The main focus of US industrial policy has been the manufacturing portion of the chip industry, but again, that can only be understood as part of a broader eco-system.

In short, it’s nice that TSMC is building advanced semiconductor manunfacturing plants in the Arizona desert. But competition in the semiconductor industry moves fast and along these multiple dimensions. Half and more of the revenue from the current factory needs to be immediately reinvested, so you are rebuilding new waves of technology on a continual basis. Competition in the semiconductor industry is about making enormous financial investments, funded across ever-shifting groups of intertwined and collaborating firms, and hoping you get it right more often than not. Meanwhile, in the competitive race of the semiconductor industry, the finish line keeps retreating ahead of you.

The post Semiconductors: Moore’s Law and Rock’s Law first appeared on Conversable Economist.