The end of Moore’s Law (and the U.S. government’s aim to keep it)


In a July summit located in San Francisco, the White House officially announced the advent of its long-awaited Electronics Resurgence Initiative (ERI), signaling the first step of a coordinated plan to return the high tech chip sector to American control.

The ERI seeks to bring together academics, military experts, and leaders of semiconductor research and manufacturing to brainstorm solutions to an engineering crisis looming on the horizon––one that stems from the impending end of Moore’s Law, and the subsequent ramifications on future semiconductor production.

What is Moore’s Law?

In 1965, Cal-Tech doctoral graduate and future Intel co-founder Gordon Moore was famously interviewed by Electronics Magazine. They asked him, an industry R&D expert at the time, to predict the future of semiconductors.

Manage your supply chain from home with Sourcengine

His prediction became immortalized as Moore’s Law: a then-radical idea that the number of transistors which can fit into an integrated circuit should double every 18-24 months.

Though he initially projected this doubling effect only into the next ten years at the time – to last from 1965 to 1975 – Moore’s Law lives on to the present day.

Since Gordon Moore’s interview, the doubling effect of Moore’s Law has been observed in photography and computer electronics. Perhaps due in part to its simplicity, as much as its accuracy, Moore’s Law has come to define for tech the seeming inevitability of progress.

The end of Moore’s Law

Chipmakers will soon be done shrinking, but they aren’t necessarily done improving. The industry is searching quickly for new methods.

In 1995, Moore himself said in a paper addressing the science of Moore’s Law that he didn’t believe conventional optics allowed for the tooling of transistors of less than 0.18 microns in size. Moore theorized that this size limitation, along with rising production costs in the computer chip industry, could one day spell an end to his law.

We can find partial corroboration of Moore’s 1995 statements today. He was correct in principle that the shrinking of computer chips should soon cease. In the final publication of the International Technology Roadmap for Semiconductors (ITRS) – led by the sole industry authority on such trends – the report authors state that chips are likely to stop shrinking by 2021.

For many, this end to shrinkage could also mark an end to Moore’s Law. The ITRS report’s authors state that the only way “more Moore” could be achieved is through the development of new chip technologies, such as 3D NAND: “Stacking of NAND devices in 3D is one way to breakthrough (sic.) the 2D scaling bottleneck… [though] the larger “cell size” requires many 3D layers to continue the cost-per-bit scaling trend.”

Now, more than ever before in the 21st century, chip production R&D is a race to the finish line.

The Electronics Resurgence Initiative

The U.S. Department of Defense’s research unit, the Defense Advanced Research Projects Agency (DARPA), sees a national security risk posed by the uncertainty of today’s semiconductor industry. Anyone could be the first to develop groundbreaking chip technology, including chip manufacturers headquartered in adversarial nations.

To address this risk, DARPA announced in April of this year its plans to spend potentially upward of $1.5 billion toward “transformative advances in electronics,” with $75 million of those funds already assigned to US-based entities.

DARPA ERI infographic ©

DARPA’s press release mentions matter-of-factly that ERI is “designed to fulfill the post-scaling predictions made by Gordon Moore…”

The entities already assigned ERI funding include research universities, laboratories, and large corporations like IBM. Through six categories of development, the program aims to encourage domestic innovation in the electronics industry, with chips under the spotlight.

ERI and the US-China feud

ERI’s accelerated funding may be viewed as another development in the ongoing US-China dispute.

There are multiple fault lines along which this geopolitical conflict runs, almost all of which extend beyond the conflicts of personality leaders Trump and Xi may hold between one another.

Taiwan Semiconductor, eponymously named after the islanded nation-state in which they are headquartered, presents a manufacturing capability for the Chinese Communist Party’s tech ambitions. They are the third-largest semiconductor company by sales in the world. Coincidentally, military pressure has increased recently on Taiwan to be annexed into the Chinese system.

In August 2016, China ordered the creation through the merger of a $2.8 billion chip manufacturer, Yangtze River Storage Technology, ostensibly to improve the domestic availability of semiconductors.

With this vital tech field dependent upon vague promises of future technological achievement, when competitors from adversarial nations are trying to be the initial product developers, the United States’ top defense research entity isn’t taking any chances.

What do you think?

The White House is investing heavily in the development of chip tech for defense purposes. Do you see a need for the US to counter China’s desire to develop in the chip industry? Is the United States even a guaranteed winner in this race, after all their investments?

Discuss your thoughts below!