Although ARM doesn’t deal with the silicon manufacturing side of computer chips, it’s the top microarchitecture company for the mobile market—the vast majority of mobile devices incorporate at least one ARM-based chip. The company is therefore keenly aware of the challenges that a declining Moore’s Law presents to the chip industry. Writing at EE Times, Rick Merritt reviewed a keynote/paper by Greg Yeric, part of ARM Research in Austin: “It’s getting harder and more costly to make chips smaller and faster, but there is still hope for advancing Moore’s Law.” Merritt quoted Yeric as saying in the International Electron Devices Meeting (IEDM) paper, “‘The semiconductor industry will need to push equivalent Moore’s Law scaling through a broadening set of fronts…in an “all-of-the-above” effort [that] will include more technology complexity, investment in technology-design optimizations, and ultimately technology-system optimizations.’”
Unfortunately for the traditionally defined Moore’s Law (the number of transistors on a wafer doubles every couple years; some definitions include at a constant cost), however, this statement is less than hopeful. It sounds a bit more like the way to continue technology scaling in the style of Moore’s Law is to look beyond Moore’s Law—that is, beyond just packing more transistors on a chip. “Yeric took a balanced view in the ongoing debate of whether or not engineers will find ways to make cheaper transistors amid the complexity. ‘We can construct a scenario in which cost reduces below [the] 28nm [level], albeit at a slower rate,’ he wrote.” Again, for the traditional view of Moore’s Law, this outlook is dubious at best.
But even if Moore’s Law (as it’s typically defined) comes to a halt, a focus on other aspects of design could still yield considerable advances, as Yeric argued. When the brute-force approach of packing more transistors on a chip at an ever falling cost loses its punch, designers will be forced to study other approaches to improving performance. The interesting question is whether these alternate avenues of development will yield the same pace of innovation that Moore’s Law has enabled.
Read more about Moore’s Law