POET Technologies Inc.

I finished watching the EC presentation again a couple hours ago. My wife and I both watched the live feed from home and, no kidding, the first thing I said after it was done was, "Well, that was... boring." I guess I had been conditioned to expect a little more by the conversations between the members here.

But on this latest viewing my take on the event made a 180 degree turn. I was absolutely amazed at the amount of information that I missed on the live viewing. When you forget about what was hyped by those invested in the company and just accept the presentation for what it was, a celebration of the achievements of Dr. Taylor and an introduction to the POET process and why it is the answer to "What's next?", then the event was a great success.

I hope that the whole presentation, slides and all, is available soon on POET's website because I think that by just listening to Dr Taylor talk about them without seeing them a lot is missed. But I was actually impressed by the detail that he went into. Just enough was expounded to explain the concepts and give reason to dig further.

I think one thing he did really well while explaining the need for POET was dispel the thought that graphene has any chance of competing with POET. He started by explaining why we must move to optical from electrical for data transmission (starting at 13:36) because as we have to move more and more data around, the physical limitations of copper dictate that as the frequency of the signal increases, the distance that you can propogate the signal decreases. (example given of 20" limit for a 18 Ghz signal) Future speeds dictate the need for optical transmission closer and closer to the source to the point where optical will be needed "chip to chip". And then he went on the differentiate POET from the current solutions, POET being the only one with an on-board laser and small size and low power requirements. So back to graphene, during the Q&A when the question on graphene was asked (47:00) he highlighted the lack of optical ability, need for huge resources to develope still, and the need for a different manufacturing process. I love the way he closed it, "...it doesn't address the optical problem, and the optical problem IS the problem. We can't go further without improved connections, and you can't improve the connections unless you can perform the optical functions using the economies of scale that you have in microfabrication. Intuition tells you that's the way it has to be."

I especially like the slide entitled "POET OE Receiver"(24:35) with the thyrister information and the diagrams of the whispering gallery waveguide. (As MrsShepherd informed us, the POET logo is taken from a thyristor circuit diagram like the one in the upper right of that slide.)

I think the answer to the "Joel's" out there that scoff at POET succeeding at 100nm was given to us at this event. Starting at 22:30 in the video Dr. Taylor gives a good comparison for distributing the clock in an optical circuit. POET's thyristor has 3 transistors, is just 20um across and performs the same function as the CMOS Amplifier which contains over a hundred transistors and measures 700um X 200um. And if I understand him correctly, 16 of these are needed, one at each of the extremities of the H pattern shown on the slide. So do the math on that and one can easily see why we don't need to be at smaller feature sizes to fit more funtionality into the same size die with drastically reduced power requirements. Interesting note on the slide that can barely be read says, "In current SOC's, clock power is 40% of chip power".

So I'd like to say congratulations to Dr. Taylor and his team in the lab, as well as to management at PTI. Looking forward to the next few weeks and months ahead.

Green