IBM recently showcased a brand new graphene computer chip, outperforming previous graphene-based chips by a factor of 10000. Apart from a tremendous speed up, it is also remarkable for the process used in its production. It is a standard CMOS production process, widely adopted in the currently dominant silicon chip manufacturing. Mapping the outstanding technological potential of graphene onto a well-developed industrial process hints at some possible future outcomes such as cheap supercomputing and superfast massive-parallelism.
The chip, born at IBM Research, is not a general purpose processor. It's not even a digital chip. What it does is receive and restore wireless signals in the frequency range of 4.3 GHz. Its true purpose is to construct graphene chips. The approach employed in previous attempts of Graphene Field-Effect Transistor (GFET) manufacture involved building the active components (transistor and graphene channel) of the chip first, followed by installing passive elements (capacitors and resistors) and connections. This didn't play very well with graphene’s delicate structure and led to many wastes in time and material.
IBM's innovation entailed building the passive elements on the silicon wafer first and then using it to capture a layer of graphene floating freely in a bath of solvent. This way they were able to obtain a fairly uniform graphene covering of all the chip structures. The downside, as the IBM engineers inform us, is the average quality of graphene thus obtained.
So what about digital general purpose chips to be used in computers? Currently the IBM team seems to be focusing on carbon nanotubes as the main contender. Its main advantage over graphene is the non-zero band gap between electron bands, which is essential for digital electronics. Graphene, with its zero gap, seems more appropriate as the material for special purpose analog coprocessors, accompanying the main CPU.
What IBM Research has shown is clearly a proof of concept, serving demonstrational purpose more than that of practical need. Pushing technological barriers and adapting known materials to new applications helps in a way shape the future of our computing and put our growing expectations in a predictable hardware context and technical perspective.
About SaMaterials: http://www.samaterials.com/
Stanford Advanced Materials (SAM) Corporation is a global supplier of a series of pure metals, alloys, ceramics and minerals such as oxides, chlorides, sulfides, oxysalts, etc. Our headquarter, located in Irvine, California, USA, was first established in 1994, starting to provide high-quality rare-earth products for research and development (R&D).