There have been an increasing number of reports lately that Moore’s Law is dead, or soon will be. Gordon Moore, one of the founders of Intel, first predicted that the number of transistors on an integrated circuit chip would double roughly every two years, bringing exponential improvements in consumer electronics. This developmental pace has continued, apparently unabated, for the more than forty years since Moore’s declaration. But experts are now beginning to question whether the increase will taper off, or if it already has stopped.
Professor of physics and popular author Michio Kaku has said that scientists will reach the upper limit of Moore’s Law by 2020 or soon thereafter. "Around 2020 or soon afterward, Moore’s law will gradually cease to hold true and Silicon Valley may slowly turn into a rust belt unless a replacement technology is found," he said. "Transistors will be so small that quantum theory or atomic physics takes over…"
Moore also projected that the amount of energy consumed by each computing element would decrease as the number of transistors increased. But according to recent claims, it appears that the limiting factor of power dissipation has overtaken transistor size reduction. This past June, a paper delivered at the 38th International Symposium on Computer Architecture reported that, "regardless of chip organization and topology, multicore scaling is power limited to a degree not widely appreciated by the computing community." The paper, co-authored by a Microsoft computer scientist, refers to "dark silicon," where the power dissipation has grown more quickly than the transistor count, requiring designers to "power down" portions of the chip so that the overall device does not succumb to heat-induced catastrophic failure. They go on to predict that advanced chips will soon need 21% of their transistors to go dark at any one time.
"Simply taking old processor architectures and scaling them won’t work anymore," said William J. Dally, chief scientist at NVidia. "Real innovation is required to make progress today." At the same time, Dally concedes, "The good news is that the old designs are really inefficient, leaving lots of room for innovation." Some scientists at Intel also have a positive outlook, and the processor giant recently announced a novel 3-dimensional architecture that increases transistor density-albeit without solving the dark silicon issue.
Is the sky falling? Can this bad news possibly add to last week’s gloomy market downturn? Are we headed for a technological slowdown? Not by a far cry, but it could continue tilting the tables away from PCs and toward tablets, something we’ve begun to see. In fact, some skeptics could say that the noise is a marketing ploy to bring more attention to low-power architectures-especially ones supporting video graphics-something desperately needed if the growth of smartphones and tablets is to continue. One example: NVidia’s latest mobile chip, the Tegra 2, is based on a power-efficient chip architecture from ARM Holdings and has started to appear in a number of smartphones like the Droid X2, and in tablets like the Samsung Galaxy Tab 10.1.
But these are just a few companies tackling what will inevitably become a big technological issue. The end of Moore’s law brings huge economic and societal ramifications at a time when we need the graphs to point up, not down. Here’s one solution: let’s call on our politicians to stop wasting our time (and tax dollars) fighting over issues that don’t help the economy or society, and put money and thinking into new jobs that will develop and grow more high-tech industries. Twenty-first century capitalism is largely technology driven; enable innovation, and the market will follow. -agc