IBM has provided an outline of its plans to invest $3 billion over the next five years. It hopes the cash injected into its R&D activities will help it find ways to make chips even smaller and more efficient and research into practical alternative materials which will prove superior to silicon.
Big Blue has observed the diminishing returns in the process reduction of silicon chips over recent years and has now decided it needs to bet big on the task of finding a suitable successor. Tom Rosamilia, Senior Vice President of IBM Systems & Technology Group said "We really do see the clock ticking on silicon". Looking about five years ahead he said "As we get into the 7 nanometer timeframe, things really begin to taper off." He reasons that processors will become unreliable when shrunk any further.
A commentator outside of IBM thought that the firm has timed its decision well, "You might say this is not a good time to be in the silicon chip business, but it is a great time to be ready for the next thing. This is the next thing," said Richard Doherty, who works as a research director at The Envisioneering Group in Seaford, New York. Doherty reminded us that Silicon chips waste a lot of their input power as heat rather than for computing purposes and "Silicon-electron mobility is like moving through snow. You can't run as fast," he explained.
The focus of IBM's research into new chip materials is expected to be upon graphene and carbon nanotubes, initially . Using these new materials it hopes to make computer chips fast, efficient and powerful enough to offer similar processing power to the human brain within the same size constraints.
IBM is also looking to exit the chip fabrication business, preferring to work simply on the IP behind processor technology. A deal with Globalfoundries Inc is rumoured to be close to completion.