As the time shortens towards the end of the fourth quarter 2014, I felt it useful to remind ourselves of why this investment is worth hanging onto. One aspect of technology investment, particularly in the area of sophisticated microprocessors, is that almost everybody undertaking due diligence has no or very little understanding of what is being offered, particularly when comparisons with existing technology cannot be made directly. This, almost certainly, means that most individuals purchasing shares now have a poor understanding of what is being claimed.

Thus, we have a 40 nM PET microprocessor which is claimed to be as faster and cooler than a Silicon based processor some 4 nodes smaller, no direct comparison of this claim has yet been published, though this may well change within a matter of weeks. I suspect there are scientists and engineers who have a very clear idea of the differences; we have certainly been led to believe there are substantial measurable differences.

Perhaps the most critical points to make are about Silicon processors. Nodal size is a term bandied around but essentially relates to the length (in nano-metres or nM for short) of the electrical control of a transistor. In Silicon this is now very small at around 14 nM. When it’s this small, crazy quantum effects creep into a transistor’s behaviour; all well described, if you can be bothered to read about it. Take it from me, it is these difficulties which have required billions to solve the reliability of transistor control. Intel has developed a 3-dimensional transistor (known in the jargon by the acronym a FINFET) to mitigate and control these quantum effects. What is certain is that billions more have to be spent to reduce the node size to the next step. It may not even be achievable but INTEL (they lead in this field) try to sound confident. When and if they achieve it they may expect a 25% improvement in processing speed. Other problems such as cooling; circuit integration, lithography (to print the circuits during manufacture); interconnects to the exterior all have formidable problems additionally. I do not believe it unreasonable to state that it will cost 10 – 20 billion of US$ in Research and Development (R&D) before they know if it can be achieved. That is not a pleasant prospect save for those charged with the development. Failure can and will result in bankruptcy

Furthermore, there are clear, unequivocal (in my view) signs that other processor developers are in distress by what is happening. There is a pervading sense of doom apparent. This distress is witnessed by plant closures, scientific murmurings about the end of Moore’s Law and the appearance of articles in learned journals contemplating the future. As I write, the buzz of distress has steadily increased over this last 12 months. Hence, there has been straw grasping with issues such as graphene, quantum computing, other semiconductor materials.

Now here’s the light. At this very point in time, an idea, developed over more than 20 years is just coming onto this scene. The POET concept is almost unknown but offers not just hope but tangible and perhaps demonstrable benefits. Back, in the early days of semiconductors, it was clearly recognised that certain elements such Gallium, Indium, Arsenic, Phosphorous and others (if you’ve done chemistry in school they are elements in that part of the periodic table known 111 – V groups) were superior to Silicon in their semiconductor properties. But, Silicon was sand cheap, and was the first developed for transistor usage. It took over but has had its day at this point in time. Scientists such as Geoff Taylor, there are not many, persisted with the 111-V elements and he has also been very conscious of the need for not just transistor performance, but of manufacturing processes, and circuit integration, so that what takes a number of separate processors can be placed on a single processor. Also, other features such as lasers, analogue and digital mixed signals, memory as well as processing can be included on the same processor. The effect has been to produce a much faster chip with substantially less power usage because of voltage reduction in the uses of these elements (there are many more available electrons with much lower resistance and light is used for information transfer within and without the processor).

So, just as its gets vastly expensive for Silicon development, in comes PET processing. You can manufacture it in existing foundries with low cost alterations. It is intrinsically faster, with higher clock speeds and, currently, at much longer gate lengths than Silicon. The gate length has the potential to be made smaller in nodal steps, so that PET processors will get faster and faster over time, passing through nodes already achieved in Silicon. Much more importantly, at this point in time, there is simply nothing else and to copy it will take some years. My take is this will address processing speed at much reduced power usage and will be a highly desirable way forward, available over a few short weeks perhaps a few months, I cannot tell. This looks and feels like a perfect storm. I say this because I can have some confidence that no man of Ajit Manocha’s reputation (he has an impressive scientific as well as business background) would take on the task of joint chairman of POET Technology without a very good reason; you don’t buy/rent offices in Silicon Valley based on failing technology and you don’t maintain silence to your investors unless and until you have a series of important announcements lined up before you move to a more investment savvy exchange and silence suggests others are very interested.

We are entering the last phase before recognition. There will come a time, shortly, when it will dawn on others what is going on. I strongly suspect that, as I write, there are a few individuals who understand this completely and know much more that I do. It will be no surprise to me if the share price will rise by some multiples when this occurs; with a very bright future beyond that over some years.

David