The challenge of designing the new generation of computer chips

Not long after Gordon Moore proposed in 1965 that the number of transistors that can be recorded on a silicon chip would doubling approximately every 18 months, critics began to predict that the era of the “Law Moore” is closer to an end. Recently, industry experts have warned that the progress of the semiconductor industry is grinding to a halt, and that the theory of Moore, cofounder of Intel, has come to an end.

If so, that will have a dramatic impact on the computer world. The innovation that has led to personal computers, music players and smartphones is directly related to the increasingly low cost of transistors, which are now stranded billions into small chips of silicon (computer chips) that can be sold for a few dollars each.

But Moore’s Law is not dead, just evolving, according to scientists and engineers more optimistic. Their argument is that you can create circuits that are closer to the scale of individual molecules by using a new class of nanomaterials (metals, ceramics, and polymeric or composite materials that can be arranged from the “bottom up”, in rather than from the top down.

For example, designers are developing semiconductor chemical processes that enable “self-assemble” circuits to make the materials to form patterns of ultra fine wires to a semiconductor wafer. The combination of these patterns of nanowires with conventional chip manufacturing techniques, scientists say, will lead to a new class of computer chips, thereby keeping alive Moore’s Law and reducing the cost of manufacturing the chips in future.

“The key is self-assembly”, said Chandrasekhar Narayan, manager of science and technology Almaden Research Center, of IBM, in San Jose, California. “We use the forces of nature to do our work for us. The Brute force doesn’t work anymore; it is necessary to work with nature and let things happen by themselves”.

To accomplish this, semiconductor manufacturers have to move from the era of silicon to what might be called the era of computer materials. The researchers in Silicon Valley are using new and powerful supercomputers to simulate their predictions and are leading the way. While semiconductor chips are no longer made here, it is likely those new classes of materials that are being developed in this field to change the computer world in the next decade. We use the forces of nature to do our work for us. Brute force does not work anymore, you have to work with nature and letting things happen on their own, says Chandrasekhar Narayan, a researcher at IBM on chip design.

“The materials are very important for our human societies”, said Shoucheng Zhang, a physicist at the Stanford University, who recently led the work of a group of researchers to design a tin alloy having superconducting properties at room temperature. “Entire epochs are named in connection with the materials: the Stone Age, Iron Age, and now we have the Age of Silicon in the past, were discovered by chance Once we have the power to predict material, I think.. which is something transformative”.

What drives this research is the economy, specifically the amazing cost semiconductor manufacturers expect to pay for their next-generation factories. In the industry this chip manufacturing called “Moore’s Second Law”.

Within two years, the new factories for the production of microprocessor chips will cost 8 to 10 billion dollars, according to a recent report by Gartner (more than double the current generation). That amount could increase to between 15 and 20 billion dollars by the end of the decade, equivalent to the gross domestic product of a small country. The exorbitant costs that will be needed soon mean that the risk of error for prime companies is immense. So instead of investing in expensive conventional technologies that can fail, researchers look for these new materials self-assembly.

In December, researchers at Sandia National Laboratory in Livermore, California, published a paper in the journal Science describes progress on a new class of materials called “metal-organic frameworks” (MOFs, for its acronym in English). These are crystalline assemblies of metal ions and organic molecules, which have been simulated with high-performance computers, and then verified experimentally.

What scientists have demonstrated is that they can create conductive thin films, which could be used in a wide range of applications, including photovoltaic energy, sensors, and electronic materials. Scientists say they now see ways to go beyond conductors, semiconductors towards creating too. According to Mark D. Allendorf, a chemist at Sandia, there is very little that can be done with conventional semiconductors to change the behavior of a material. With MOF, imagines a future in which the molecules can be sorted with precision to create materials with specific behaviors.

In November, scientists at the SLAC National Accelerator Laboratory, in their article for the journal Physical Review Letters, described a new form of tin in a single molecule thick, it has been predicted that can conduct electricity with an efficiency of 100 percent at room temperature. Until now, it has only been found this kind of efficiency in the popular materials such as superconductors, and only at temperatures close to absolute zero.

The material is an example of a new class of materials called “topological insulators” which are highly conductive along a surface or edge, but insulated inside. In this case, researchers have proposed a structure with atoms added to a single layer of tin fluorine atoms. The scientists, led by Dr. Zhang, staneno named the new material combining the Latin name of tin (stannum) suffix used to graphene, other material based on a sheet of carbon atoms of one molecule thickness.

The promise of such material is that it can be easily used in combination with current chip manufacturing processes to increase the speed and reduce the power consumption of future generations of semiconductors. The theoretical prediction of the material has yet to be verified, and Dr. Zhang said the investigation is being carried out in Germany and China as well as in a laboratory at UCLA (University of California, Los Angeles).

It is quite possible that the revolution in computer hardware can provide a cheaper way to the next generation of computer chip technology. That’s the bet IBM. The company now experimenting with exotic polymers is automatically formed in an ultrathin network and that can be used to form circuit patterns on silicon wafers. Dr. Narayan is cautiously optimistic, he says there is a good chance that self-assembly techniques from the “bottom up” eliminate the need to invest in new lithographic machines, which cost $ 500 million, and that use X-rays to record smaller circuits.

“The answer is probably yes”, he said, describing a least cost route toward denser computer chips.