Is Moore’s Law about to run its course? Whether or not the anticipated doubling of computing power every eighteen months finally comes to an end, the long-term consequences for military technology will be profound. And in the prediction business in the past two months, we have seen diametrically opposed views.
In the World in Review 2014, Benjamin Sutherland, freelance correspondent to The Economist, inquired “No Moore? A golden rule of microchips appears to be coming to an end.” The magazine argued that while manufacturers “will probably manage to cut transistor size in half at least a couple more times… beyond 2014 shrinkages will no longer cut transistors’ cost.” While this still may be useful for the most demanding applications, military buyers with more price-sensitive needs will not be excited.
In contrast, Harald Bauer, Jan Veira, and Florian Weig of McKinsey & Company asked last month whether Moore’s Law is up for “Repeal or Renewal”, and concluded that it has some time. Extreme ultraviolet lithography, multicore system-on-a-chip architectures, and yet larger-scale manufacturing may help extend the trend in the short to mid-term. In the long term, they look to carbon nanotubes, quantum computing, and even spintronics as bases of technical advance.
Even if computing power stops advancing soon, the state of the art in software will continue catching up for some time. Software developers have long been able to take advantage of expected advances in processing power to write slightly sloppy code quickly. They could always rely on advances in hardware to ameliorate much of the executional problems.
But at some point, an end to Moore’s Law would mean an end to endlessly amazing, decade-on-decade advances in C4ISR systems. Consider robotics, but without cheap advances in computing hardware. Dire predictions of how the lethal autonomous agents of Hebollah will soon be wandering the planet would then, in a few years, seem as silly as personal jetpacks and freeze-dried ice cream.
On the other hand, if quantum computing or some other technology—today thought exotic—saves the geeks from boredom, other pillars of military advantage may being to fall. Consider air defenses, which in several wars around the world have given NATO’s air armadas little trouble. Stealthiness is mostly baked into an aircraft when the design is finalized, and convincingly lowering observability in a retrofit is challenging (though Boeing is trying with its Silent Eagle). At the same time, sensor technology advances just behind Moore’s Law; if this continues, even the vaunted F-35 may become “not that complicated” to spot.
In short, the state of the advance of computer hardware technologies will have a huge effect on not just technical architectures, but even what is desirable and possible in national strategy. This is an issue that strategists in industry and the military need to watch, and consider carefully when crafting narratives about the future of warfare.
James Hasik is a senior fellow in the Brent Scowcroft Center on International Security.