Graphene is, among other applications, being regarded as a potential replacement material for silicon, which has begun to show its age in face of extremely rapid technological progress, states a recent article written at McKinsey & Company. Silicon, the primary material used in the semiconductor industry, has historically kept pace with Moore’s law by providing previously unimaginable progress. However, performance improvements are raising the cost of silicon technology, as are slowing performance improvements, while miniaturization is reaching the limits of this material. As silicon appears to be reaching the final plateau on its s-curve, the industry is experimenting with several potential replacement materials, of which graphene is regarded as the most promising one.
There are three leading indicators that silicon technology is reaching its developmental end. First, the performance advances of computers have slowed down considerably. Starting in the 1970s up to the 1990s, computer performance advanced exponentially, according to the McKinsey report. Beyond 1990s and until 2005, computer power was advancing by the leaps. After that time, however, a very noticeable slowdown in progress is more than evident. Similarly, smartphone processor performance advanced rapidly until the year 2013, after which we are witnessing a plateau. Such slowdown affects companies that built their competitive advantages on continued innovation, which are now seeing their lead eroding. As competitors begin to enter the market, prices necessarily fall, which puts financial pressure on market leaders. Such leaders will now be forced to innovate, looking into ways to keep improving device performance.
The second indicator for the end of the silicon era is the increased cost related to manufacturing smaller electronic circuits. Moving to 14nm resolution several years ago increased capital and R&D costs for fab. The next step (transitioning to 11nm) which is required for faster chips is estimated to increase capital costs by 40%, while R&D costs are expected to rise by 150%. Manufacturing equipment is the dominant capital cost factor, because such high resolutions have surpassed the physical limits of electron beam lithography which has now moved to multipatterning to overcome this barrier.
The third and final indicator is the physical limitation imposed by the small size of today’s transistors. The transistor channels are now so small that quantum effects are starting to play a role, causing noise, leakage, and heat issues. Even though companies that rely on silicon technology must keep investing in performance improvements, given these three looming limitations they are forced to shift their investment towards new materials.
Graphene and other 2D materials hold obvious potential for future-gen electronics beyond silicon. A most striking advantage is their thickness of just one atom or one molecule, which in itself provides advantages over silicon. Key disadvantages of graphene, according to McKinsey, are the long R&D road ahead to commercialization and the risk of unknown hurdles at scale-up to industrial production.
The report identifies two technical and two industrial limitations for the adoption of graphene. The first and most obvious limitation is the absence of a band gap in graphene, which results in the inability to switch transistors off. The second technical limitation is high-quality production on a large scale. The graphene community is aware of both of these limitations and has been working hard to address them, resulting in for example 6” graphene wafers. On the industrial side, one issue is that semiconductor companies have already tied their large investment plans to improving existing fabs, so it will take courage to invest in graphene-based fab. The second industrial issue is that the entire integrated value chain needs to be reworked to include graphene, whereas silicon value chains are well established.
Given these challenges, it comes natural to predict that graphene adoption and market growth will come in the following phases: graphene as an enhancer of existing technology, silicon replacement, and finally revolutionary electronics. In the imminent enhancement phase, graphene will provide incremental performance improvement to cutting-edge devices, such as interconnect speed, transistor lifetime, and heat conduction. McKinsey predicts this phase to last for another 10 years, beyond which graphene will replace silicon, leading to improved feature size for next-gen CPUs and memory. Beyond that, phase three that is predicted to commence at 25+ years, should lead to new and undiscovered applications, perhaps based on spintronics, valleytronics or other quantum computing platforms.
If all goes according to plan, the report predicts a market of about $70 billion for graphene semiconductors by 2030. It is certain that large multinational technology companies are already investing in graphene, and that at least some of them will seriously consider graphene as a potential silicon replacement material. The McKinsey report suggests an “innovation X-ray”, a series of 10 questions for technology executives to consider before making the decision on investing in graphene research. Answering those questions should lead to a “structured innovation” strategy, enabling businesses to participate in these exciting new developments while minimizing the risks.