“Moore’s Law,” the famous prognostication coined in 1965 by Intel co-founder Gordon Moore, predicted that computing power would double every 18-24 months. The Law has more or less been proven true for decades but has recently been widely reported to be reaching the end of the line soon, due to the simple fact that transistors have gotten smaller and smaller and have become too densely packed.
Many experts have gone on record to say that engineers are reaching physical limitations of power consumption and cooling, including Bob Colwell, Intel’s former chief architect stating that by 2020 Moore’s law will cease to be a viable paradigm, with a bold prediction that its passing will also have significant repercussions to the continuing development of the information economy.
That would be 50 years in a row of computing power doubling every two years more or less like clockwork suddenly screeching to a halt. Consumers accustomed to ever-increasing sophistication of computers, software and mobile devices would be left for the first time wondering what to do now that computers have reached a perceived physical limit.
Enter the “NanoFSM,” developed by a team from Harvard University and the non-profit military contractor The MITRE Corporation.
NanoFSM is an ultra-small, ultra-low-power processor—termed a nanoelectronic finite-state machine that is smaller than a single human nerve cell. It is composed of hundreds of nanowire transistors, each of which is a switch about ten-thousand times thinner than a human hair. The nanowire transistors use very little power because they are “nonvolatile.” That is, the switches remember whether they are on or off, even when no power is supplied to them – which has significant advantages in terms of heat dissipation. The team behind it believes it to be “the densest nanoelectronic system ever built” and that looks like it might allow Moore’s law to continue onward, just not under Intel’s stewardship any longer.
Others working to stave off the end of Moore’s Law at the University of California at Berkeley are working with nanoribbon graphene, a single layer of carbon atoms, just one layer thick that researchers are promising could lead to transistor densities on a computer chip as much as 10,000 times higher than what is achievable today, aided in large part by the unique material’s superconductivity.
Making predictions is a dangerous business, but just as the technology we take for granted today would seem like magic or science fiction a few short years ago, there are still brilliant minds working on maintaining the pace of innovation, and pushing it to new heights unimaginable by consumers today. The real lesson here is that any time someone tells you something is not possible, in Hosting, Computing, Cloud Software, Business, or Life… remind them, it’s only not possible… yet!