A few months ago Gordon Moore stated that More Law is dead. It is nearly 40 years since the publication of his law which states the transistor density on integrated circuits double about every two years. This is quite obvious as at one point one will find we cannot build a transistor with just a small number of atoms. This was originally listed at Salar Golestanian Blogs
Gordon Moore is now retired from Intel and when finally, asked if there were any new laws for next 40 years, he said: "I'll rest on my laurels on this one! I'm not close enough now to make new predictions - several things have been called Moore's Second Law but I can't take credit for any of them."
Gordon Moore's warnings are backed up by a new report from the US National Research Council titled The Future of Computing Performance: Game over or next level? It's bad news on many levels for a trillion-dollar industry that has become an engine of economic growth.
The report states that one of the main limiting growths in computing performance is Power consumption. This is because as transistors have become ever smaller and more tightly packed, the speed at which microchips are clocked has levelled off, reaching around 3 gigahertz in 2005. That's because such fast chips generate too much heat to be used in SmartPhones and the new trends in products like iPad or very thin small laptops that seem to sell well these days.
One approach in getting faster speeds has been to fabricate two, four, or eight -microprocessor cores on a single chip. This trend has its ceiling to be reached as the power efficiency of present transistors cannot be improved much more.
To get round this problem we have to develop new transistor architecture. as well as improvements in software to allow tasks to be shared in parallel rather than serially. This idea is not new as IBM Cray Research work was done some 30 years ago.
The wall
So it seems that we’ve hit a technology wall. We can still double the number of transistors on a chip every couple of years. We can still double disk drive capacity every 2 to 3 years. We can build faster interconnects, such as QuickPath, Light Peak and 10 GB Ethernet. However, the easy improvements are over. Going forward performance gains will be measured in single digit percent each year.
The Implications to Information technology - like most of the western world, is driven by consumer spending. So what happens when a new PC is only 10-20% faster than your fully paid for three-year-old PC? Today, we can no longer count on performance increases to open up new application territory. Therefore we will see differentiation move to what were once considered secondary characteristics.
For example in the past 15 years we have seen great achievements in power efficiency in the mobile space. And in some cool devices like iPad or the new MacBook Air the size of the PC board is so thin and we don’t see many chips on them. Just a few chips and a big battery dictating long battery life is a must now
Added functionalities are also what drives sales, where adding a high end camera in a mobile phone seems to do the job.
The only good news for this the IT world is that the Data Storage industry is continuing to grow and double storage every couple of years at the same price point.
Here are some examples of what kind of technology could be a game changer and let Moore's Law continue for another 20 years.
Researchers harness chaos theory for new class of CPUs
Ultra thin Alternative to Silicon for Future Electronics