One comment of made about the netbooks, ultraportables and other small computer is that they will eventually become irrelevant as hardware becomes smaller, less expensive, and more powerful. A modern netbook has as much processing power as a high-end gaming or workstation computer from ten years ago. Given that rate of advancement it seems reasonable to believe that we will in ten years have phones with power equivalent to a modern notebook. If that were to happen low-cost computers would become as numerous as bananas and virtually all consumer computers would be of netbook size or smaller (if you needed a larger display you’d simple plug one in).
Intel’s latest post in the IDF blog offers some insight into how this process of miniaturization occurs and points out the technological advances that have made it possible to produce smaller and smaller processors.
Intel has been introducing new generations of process technology on time and on a 2-year cadence for many generations. The first 32nm products started shipping exactly 2 years after the first 45nm products, before any others in our industry, and the 22nm technology is on track for production readiness in late 2011. Intel’s research group has a variety of novel transistor and interconnect ideas in the “pipeline” including III-V channel materials, multi-gate transistors, 3-D stacking and others.
The blog post essentially concludes that Moore’s law will continue to work as expected allowing transistor counts to increase exponentially. Intel will be releasing parts based on a 22nm process in 2011 and given the typical 2-year cycle at Intel the next generation of process would be finished in 2013. But it also goes without saying that Intel is invested in this being the case. Manufacturing technology is one of Intel’s largest strengths. Any interruption in Intel’s progression towards better manufacturing technology would be a blow to the company.
There have been rumors brewing that Intel’s happy go lucky approach towards Moore’s Law is simply self-serving marketing and not engineering reality. has stated that Moore’s Law may soon no longer apply, and even Gordon Moore himself has said that he believes we’re reaching a fundamental barrier (the atom, and I don’t mean the Intel processor) which will make new manufacturing processes harder and harder to implement until, finally, it will no longer be possible to down-size them.
This doesn’t mean that the number of transistors in processors won’t go up, but it does mean that it will harder to increase transistor counts while keeping processors small and power-efficient. And so we probably wouldn’t be seeing netbook sized computers as powerful as today’s fastest machines.
Of course, this is a lot of speculation. Intel seems to have been hinting that they will find ways to continue Moore’s law, the size of the atom be damned. There are also of course more extreme fields of research, such as nanotechnology and quantum mechanics, which could come into play. But my bet is that we will run into a barrier which will make it much harder to fit powerful processors into tiny PCs without dramatically increasing costs.
You must log in to post a comment.