In the battle to keep laws of physics at bay for another 18 months, keeping Moore’s Law ticking over research groups from around the world have been locked in a race to produce silicene. A race that appears to have ended in a four way draw, four teams from around the world have succeeded in growing silicene on a silver substrate.
Silicene is a single atom thick layer of silicon that scientists hope to use to make our gadgets of the future. A wonder material of the future it is one of the most conductive materials known to man. In the same class of nano-materials as graphene these highly conductive materials see atoms travel upon their surface like light-speed ice-skaters.
A French and German team have now published their results, with initial testing demonstrating they have produced a single atom layer of silicon on a silver substrate. The production process not only uses common ingredients such as silicon and silver but it also relies on standard production processes, showing promise for easy integration into current chip making equipment.
Theoretically, silicene should allow chip speeds similar to the now infamously fast, and difficult to work with, graphene. IBM’s experiments suggest chips made on a graphene base, single layer of carbon, are capable of Terahertz speeds.
Silicene is the most promising technology for the continued advancement of chips in the not too distant future. There is often a long lead time for such new technology to actually being installed in the semiconductor factories of the future, silicene may surprise many. If silicene promised only an incremental improvement in performance that lead time could be decades but silicene promises a generational change. With huge performance improvements possible and little risk or change to process required silicene is the rising star of the semiconductor world.
Another tool keeping our insatiable need for computing speed fulfilled, Moore continues to be Law, while the laws of the nano world become a little clearer with each new discovery.
With incredibly poor timing Michio Kaku this week released a video discussing the possibility of Moore’s law collapsing in 10 years.
“If I were to put money on the table, I would say that in the next ten years we’ll simply tweak Moore’s Law a bit with chip-like computers in three dimensions, but beyond that we may have to go to molecular computers and perhaps late in the 21st century quantum computers.” – Kaku
Every 10 years or so there seems to be a declaration that the sky is falling, there is no way we can keep advancing at this pace they cry. While Kaku claims that 5 atoms thick is as low as we can go with silicon, the silicene discovery quickly moves that limitation back to one atom. Many commentators Kaku included seem to miss the point of Moore’s Law, it isn’t so much a measure of technologic prowess, it’s a measure of human creativity.
For clarity’s sake Kaku is obviously a Moore’s law purist. Moore’s original law says transistor count will double roughly every two years. Later modified by Intel’s David House to stipulate that processing power will double every 18 months, this is the more commonly accepted variation of the law. Other variations include programming complexity will double every two years until programmers heads begin to explodes.
For the popular Moore’s Laws – processing power – to continue, shrinking and upping Mhz aren’t the only tools available to chip designers. There have been a number of periods in the past 2 decades that have already demonstrated this. In the great race between AMD and Intel in 1999 both companies had extreme difficulty getting their chip designs past the 1Ghz barrier.
With CPU speeds barely increasing for over 12 months the Pentium 3 and AMD K6 processors had indeed hit the wall. Many feared Moore’s Law would fail for the first time ever. AMD won the battle past the 1Ghz mark – x86 CPU- in the end by integrating the memory controller while doubling processing power by going wider with the first dual core x86 CPU.
More importantly that generational change was done with very little process change but instead the improvements were in the design of the chips. It was the last time Intel tried to do design and process change together, with a series of fumbles that eventually saw the birth of the Pentium 4 CPU. Intel switched to the tick tock system of development.
There are many strings in the bows of engineers and scientists when it comes to improving our computing lives. While the manufacturing process or node is the most commonly discussed, the design of the chips and the fundamental physics involved are also important design considerations.
Moore’s Law is in many ways just an interesting result of our battle with the laws of physics, our struggle to understand their true nature. Our ability to continually improve the power and efficiency of our semiconductors is based on the fact that we still have much to learn about the world of the nano sizes. If we were even close to approaching the limits of knowledge in that field then Moore’s law may well be in trouble, but thankfully there is still much to learn and interesting advancements to be made, in the world of the nano especially.
Related Article: New Scientist