IBM has been developing cutting-edge technology since the early 1880s. The company has designed everything from electric tabulating machines (a fancy name for calculators) to the Power8 12-core super processor packed with 4.2 billion transistors.
"Billion" seems to be the word of the year for the technology powerhouse, which recently announced it will dedicate $3 billion in funding over the next five years to research and development of new processor technologies.
Moore's Law states that processor speeds or processing power doubles every two years; both speed and transistor count will eventually reach a limit as transistors continue to shrink. The theory is rapidly becoming true. There is only so much researchers can do with silicon as the processor's manufacturing medium. IBM is looking for microchip solutions that do not use this element. It wants to use materials such as graphene and carbon nanotubes to allow the chips to become smaller and faster.
12 CPU cores on a single die makes up IBM's Power8 processor.
Using the new materials will allow researchers to scale the processor's transistor size down below the 7-nm limit imposed by silicon and therefore allow the transistor count to be increased on the die. The company plans to use a portion of that money to work on neurosynaptic chips that emulate the human brain's computing and power efficiency. This will eventually lead to neurosynaptic systems (super computers) that feature billions of neurons and its own programming language to execute programs while consuming only a kilowatt of power and fitting into a space no bigger than a 2-liter bottle of soda.
Quantum computing is also on the table. This system uses "qubits" -- the polarization of a single photon -- instead of transistors to compute information. Zeros and ones are on the way out.
Lastly, IBM will use part of the R&D money to look at ways of furthering the use of silicon in microchips (waste not, want not). Specifically, it hopes to develop nanotech components for silicon chips designed for use in big data and cloud services. This includes the integration of Si and Si/Ge FinFETs or non-planar double-gate transistors that feature less energy bleed and electrical control than traditional transistors.
These new processor technologies could be used in a wide variety of beneficial applications, including medical research, weather modeling, and intelligent cloud services. All in all, $3 billion may seem like quite an investment (and it is for most of us), but the company spent $6.2 billion last year on R&D projects for mobile computing, Power7+ chips, and new storage-class memory for large servers. Obviously, IBM's research and development for these new technologies will take some time, but the company hopes to make significant breakthroughs in the next decade. When new products featuring them will hit the market is anybody's guess.