TM: So we've structured today's meeting into three sections. Suresh if you could advance to the agenda. First Suresh we'll provide some updates supported by recent additions to our investor deck and a preview of what poet and SPX will be exhibiting at the CIOE in Shenzhen china, the largest of all the optoelectronic shows, we had expected to preview the live demos we planned for CIOE in a press release, but the show has been postponed due to the two-week lockdown of Shenzhen. We don't have a firm date for the replacement show or the delayed show. The second part is that Kevin will review some of the mechanics of warrant conversions a new transfer agent computer share is very fussy about all the forms being filled out correctly and the process takes longer than you might expect even though the warrants will only begin to expire in April of next year. We're here today to ask each of you many of whom are our largest shareholders to exercise the warrants early and to hold and not sell the underlying shares. This is something that Suresh and I have done recently, and other insiders are in the process of doing. We are asking you to follow our lead in order to help us lengthen our cash runway since the cash from the warrant conversions comes directly to the company. We need to extend our runway for two reasons. First, to give us time to complete some agreements and projects with companies that we hope to be able to announce into the fall and later in the year that we believe will represent important developments for the company and second to give our shares and the overall market time to recover from the downdraft we have suffered these past several months. The third section is a Q&A. I'll turn it over to Suresh to begin the presentation (slide 1-6; 00:00-02:15).

SV: Hi everybody on the call. Thanks for taking the time and I know it's six o'clock in the evening for most of you on the east coast. What I intend to do today is to give you kind of a status update on where we are with our product development but really you know underline my firm belief that you know we are developing something very special. It takes time and you know we need patience, but you guys have all stood by me so far as I've tried to transform this company into delivering something meaningful and of high value and high potential and you know more so now than ever. I'm convinced we're on the right path. Even since the Brussels conference we get so many inbounds on a variety of projects and we just can't take all of them, unless they are directly applicable to what we're doing. So we're parking them till we deliver on what I think are meaningful achievements over the course of the next couple of quarters. What it proves to me is that what we're developing is of universal interest and value and that's what I would spend a bit of time talking to you about. So you know the key challenge in photonics, the key challenges are volume. Can you deliver in the high volumes required not only for today's applications but for applications in the future? Form factor or size is becoming critically important because these modules are fixed in size and as you keep increasing frequency, you have to fit more and more stuff inside of that fixed size and without integration that becomes impossible to do and then POET has a distinct advantage in terms of integration and size since everything we do as wafer-scale manufacturing. You heard me say this as well, when we ramp production, we have a disruptive cost advantage so in the assembly piece as well as power consumption. Some of the things that we're doing on the interposer dramatically reduce power consumption, because of the size. I mean, when the size is small, the distances that electrons need to travel is small and that consumes lower power, so these are the key challenges in photonics and current technologies do not address universally all of these challenges, and we do. Over the past couple of years we've added so much capability into our interposer. Of course we've faced challenges along the way, but you know, we've also overcome all of them, and I'm very confident that we're on a path here to being able to know provide, basically meet my vision and mission for the company you know which is uh which is dear to me right, in terms of what we want to do (slide 7; 02:16-05:53).

So if you look at the data communications roadmap, it keeps moving forward right. When I first joined POET it was 2015. Early in 2016, 40G was the highest volume product. 100G has just started ramping. This is 2016… so about six years ago…you know today 100G is the highest volume product. 400G has already kind of you know moved into production, but when you keep ratcheting up the speed, current conventional technologies have a hard time. Beyond 400 and even at 400 they're you know not as efficient as what we could do with our interposer. But as with any technology I mean there are incumbent solutions and there are incumbent solutions at 400 as well but when you get to 800G, then you look forward to 1.6T and beyond, the roadmap ends, and integration is needed. I think you know for those of you who follow what happens in the world of CMOS, there was this planar technology. Planar CMOS kept going on eight quarter micron three and so on and so forth, then you reached a point where planar CMOS the road map was dead. I mean, if you didn't do something different, you didn't proceed and then 3d technologies came online and then the roadmap continues. In the optic space we're kind of in that point where conventional technologies have taken it up to 400G. In fact even at 800 you can see conventional technologies that are out there today, but they're done in such a complex manner that it is hard for them to scale it up. So at 800G there is a discontinuity where conventional road map ends, and some form of integration is required, and POET is really well poised to provide that integration. Now of course you know in in hindsight you could say well why did we not shoot for 800G right off the get-go but it's tough. I mean we're developing something completely new. We had to validate it with current solutions which we did successfully at 100G, 200G and now with all that experience behind us, I feel supremely confident that we're going to be successful here at 800G. POET can support pluggable form factors all the way up to 3.2 terabits per second where conventional solutions struggle to deliver 800G (slide 08; 05:54-08:47).

Now, we're not the only ones who say this. So this is a slide from a presentation from one of the top 10 module companies in the world and they say we have reached the limits of discrete sub-assembly. We need photonic integrated circuits basically. You know, the channel counts have increased so dramatically. Eight channels, then 16 channels. The complexity of these sub-assemblies has increased dramatically and what people want is an integrated solution that just replaces all this nonsense right with the chip scale form factor and that is exactly what we're doing right now. So we're having some very interesting dialogues now because for the first time perhaps the industry has acknowledged that. You know we got to make a transition and we're able to have these conversations on the back of the progress that we've made over the past couple of years and improving and demonstrating that our solutions are in fact scalable and they do work, and they do meet the requirements of the market. All of that was important steppingstones for us to get to this point (slide 9; 08:48-10:04; no slide 10 presented).

Now, data communications has progressed systematically through increasing channel codes and that's where multiplexing and these other forms of components that we incorporate in the engine comes in handy. At the 10 gigabit per second range there was only one lane, 40G it went from one to four, then at 100G it stayed at four. Fine, current technologies you know even though they may be klugey they work. To get to 800G suddenly it's eight lanes, to get to 1.6T it's 8 to 16 lanes, 3.2T it’s 16 lanes. So it's very clear that once you get to about the eight channels required especially in a multiplexed solution, it is not possible to make these devices. This particular company they do it today I mean they do have samples out, but they are also very open and saying you know we can do, you know, we can put samples out, but try scaling it up in terms of volume… this kind of complex assemblies don't work. In fact, these assemblies from many customers or potential customers use up to 50 individual pieces that they put together. You can imagine that you could, you know, these are engineering feats of marvel, they're actually able to do it, but can you actually scale it? Can you get to the cost structure that you need? Those are the issues and we have just completed, happy to say, a design for the 800G. That looks like what I show on the picture below. It's 75 smaller. You fit the receiver and transmitter side by side, right next to the DSP, and most of the module is actually empty. You can see readily how a solution like this not just addresses kind of the cost, scale, volume, but the proximity to the DSP means also better power. All this empty space in the module means better thermals and better manufacturability right, and so we we are extremely bullish about our ability to really make a splash here. I mean, people have been paying notice to POET, I mean, in most of the presentations I see now, we are a prominent potential supplier of integrated photonic integrated circuits, so we've now built a name. People are watching us, and I think this 800G is really going to be the linchpin right, in terms of being able to really swing that pendulum. I'm more confident today than I've ever been that this is the path and everything that we've worked for so far has brought us kind of to this culmination right (slide 11; 10:05-13:23; no slide 12 presented).

So we look at key competitive differentiators and we compare what we do at POET to what other vendors do and these are actually vendors of silicon photonic solutions. They include the enthals of the world and the marvals and so on and so forth and we competitively compare what it is that we could do with what others offer in their quote-unquote silicon photonic solutions. And you can see that you know when it comes to a full-featured set of solutions that are truly integrated, chip scale, we're the only ones who do it and that's because we've chosen this interposer path right. Which is a hybrid integration path that allows us to truly put all the pieces that are necessary in a chip scale form factor as opposed to just the transmitter or just the receiver or just the modulator as some folks do. So there's a lot of hype in silicon photonics and for sure and I'm not trying to compete with them. They're a component supplier, we're an integrator, but what I would say is that most of those solutions are incomplete and they don't quite get to where you need to be. And we've consciously engineered what we need to do being able to com provide a complete solution that applies not just to data communications but other market verticals as well. We've got inbounds on sensing we've got inbounds and some space applications that I'm having conversations about. People generally like this concept right and where it is going to. You know we need to complete the 800G. It is critically important for us to do that and I'm supremely confident we can, before we take some of these other applications on as well (slide 13; 13:24-15:27).

So we have a very strong value proposition at 400G and beyond, but my sights are right now within the company set on 800G. The strategy is: deliver 800G which is a two by 400G solution, show that we could actually fit it in this form factor that is insanely competitive and then go two ways we go backwards by value engineering 400G, because 400G has been in production for a couple of years. So people over the life cycle would like to value engineer and get to a lower cost solution and then propel forward um to 1.6T and 3.2T. It's important that we plant a flag right we want to plant a flag at 1.6T and say hey we could do it in fact at the CIOE we were going to demonstrate, and we will when the CIOE becomes a full 1.6T receiver. I want a leapfrog right and I want a leapfrog from, you know, a robust base at 800G. 800G is an inflection point for rapid growth in the data centers. It also extends this preferred path of pluggable optics there's a lot of debate in the industry on co-packaged optics versus pluggable to the extent that you can keep doing pluggables. It is the preferred form factor, and we believe we can provide a roadmap extension definitely to 1.6T and potentially even 3.2T and our technology surpasses all others of this node in terms of size, power, cost, ability to scale, design flexibility. If you take up so little room on the module the module design becomes very easy, which is why we're also contemplating perhaps just do in the modules. You know, all of that requires time and investment and we're being very cautious with how we spend our money, but surely, we need to put more wood behind these arrows. I’ve been very vocal this past six to nine months uh more so than I have before, because that's me you know. I was conservative to begin with I wanted to make sure that what I said had credibility and weight. And now I'm being very vocal um I'm going to be at Basel at the e-cock conference in a couple of weeks I've been invited to give a talk in March at the EPIC conference. So I think people like what we're doing and want to hear more about it (slide 14; 15:28-18:13).

So three things with the interposer we want to do one is not just enable the solution for 8G to 3.2T. We're going to do that using our optical interposer platform with the directly modulated lasers. We're very close to securing a supply for these lasers and we will likely be in a position to announce something there (slide 15; 18:14-18:41).

The second thing is the light sources for high bandwidth communications light sources for high performance computing. This is the future of computing the next generation computing AI/ML intelligence accelerators. We can reduce the cost of these light sources by 75 percent. We've got a win with celestial as you already know and we're chasing other companies um like Ayar Labs, Luminous, etc. Many conversations with some of these folks to, you know, get to being a solution supplier to them for remote light sources. So the interposer platform again for these sources that require 4 to 8 to 12 to 16 lasers in a tiny form factor that's going to fit next to your graphics card on a PCI express board. You know, you can't use clunky assembly techniques right. And so there's a there's a definite market for us here and this is, you could say you know AI and ML people it's futuristic, but really the business is there in 2024 and beyond, and its big business. This is now a one-to-one attach rate of photonics to a silicon chip right. Imagine if every microprocessor you had in the world needed a silicon of photonics chip for it to function. That's what these remote lasers are. It's like dram is to a microprocessor a remote laser is to a AI ML accelerator in the photonic space so these are large volume businesses and we believe we have um the right solution. So these are our two primary focus areas today and this is what is consuming my team and my time. We’re working very hard to get this over the finish line and into production next year (slide 16; 18:42-20:54).

The third tenet with the interposer is integrating new materials. We know that there are new materials on the horizon that are even better than silicon photonics like lithium niobate. We've had an engagement with a company there called Liobate. In fact at the CIOE, we're going to be demonstrating a 400 gigabit lithium niobate transmit solution. So the interposer is a critical component in allowing these new material systems to kind of shine right. Whereas by themselves they have no integration potential at all right. And the interposer with its hybrid integration allows for that roadmap extension with new materials as well. So those are the three things that we're working on in the company right. Make the solution for 800G, which then allows a value-engineered 400G and flagship 1.6T light sources for AI ML. Big market, huge potential, ramps a little later, 2024, 2025, but in 2025 it's millions of units a year at a couple hundred bucks right. So you can do the math. That's big business and we have to succeed there. So it's, and I know you know we said oh wow we're 2022 now, that's 2024, that's two years…but two years flies in no time especially when we're talking about needing to prepare ourselves to deliver that kind of volume. And then new material plays (slide 17; 20:55-22:41; no slide 18).

So we are aiming at these two critical market inflections points they really take off in 2024. We want to get samples out on these. By first quarter of 2023 I want to be at 400G and 800G. So that we can ride that wave. I'm trying to shoot ahead with 1.6T today, so that we can be a flagship provider as opposed to being late and slow. And I think one of the things we've done at the base interposer platform development is it's allowed us to, you know - if the components are ready - we're able to put a solution out. So now with this secured supply of DML lasers that can take us all the way to 800G. We're very, very confident that we can get there (slide 19; 22:42-23:34).

So our go-to-market plan so this is a chart. I mean any of you who have done any kind of volume business you know this right, but when you start the market, you're at market pricing. Then as you go into production, the price drops. The cost drops and you start making money. So in the early stages, it's market pricing. Whatever you charge, they pay. The volumes are really low. After that, your cost is dominated by your bill of materials and then finally your cost is dominated by manufacturing. In fact in this blue period is where people make most of their money which is cost of manufacturing. So on this curve you know 100G, 200G is on the extreme right side. I told you, it's about four or five years in production, so it's already at a fairly low cost point. We're using this to prove out platform to prove out the technology, get customers engaged. Even if they're not interested in 200G they see it they kick the tires. We're getting market acceptance and it's a tech throughout vehicle for us. We do have customers on 100G LR4. LR4 has another 10 years of life, so we are excited about that. But for FR4 or CWDM we have customers. it's going to be a small volume, but it's a really critical technology throughout vehicle for us. From there we're going to leapfrog to 800G. And that's because 800G is now in this yellow phase where it's just about getting started. No really good solution for 800 FR4 in the market yet. And our reduction in bill of materials as well as this disruptive manufacturing advantage. 400G is in the blue right now, but there are enough customers who would value engineer, because there are many, many years of production in which they can recoup their cost of retooling and make their margins due to the disruptive manufacturing costs we offer. And then finally 1.6T is a game changer. It's where it's a flagship, and I just finished a design today on the 1.6T product. The receiver is going to tape out to Siltera this weekend. We're already demonstrating the 1.6T at CIOE. This will be a production part that will go to customers and we will be demonstrating it at the OFC conference as well. So I'm trying to leapfrog here from 800G to 1.6T, to plant that flag that says hey guys everybody else is struggling at 800G and we can readily scale to 1.6T right. That's a statement and I'm confident we can do that. Our manufacturing value-add through the interposer, you know we said this before, wafer scale assembly and test, no active alignment, fiber attached capability, low loss, eliminate all the wire bonds we have through silicon vias and our interposer that can go to the DSP directly. It's a complete solution with integrated electronics. We're working with MACOM on the drivers, we're working with MACOM and the TIA’s and a couple of other providers, and we are basically incorporating their electronic components onto our other interposers. So it's a complete solution. When we talk 800G and 1.6T it's a complete solution (slide 20; 22:35-27:17).

So our go to market plan, you know, we're now sampling and selling to module customers through super photonics. That's the extreme left. This is 100G 200G solutions, also 400G receivers. We're actively in development on the 400G and 800G and we expect those to sample in 2023. And as you can see the ramp starts second half of 2023 into 2024, and then there we have basically the roadmap extension from 1.6T all the way to 6.4T. And at 6.4T we did a design with a three-dimensional stacked transceiver. Nobody's ever seen that before. We've actually stacked the receivers on top of the transmitters to pack that density. I mean, the capabilities are endless. We’re limited by our own bandwidth. We're limited by our own ability to invest. I mean I'm doing a 1.6T transmit and receive at a million dollars of R&D spent a month. I mean and I'm competing with Innolights and Finisars and, you know, billion dollar companies right. So that goes to show that what we have is simple. It can be done. And we're poised to do it now. And you guys have supported us over the years. I know you've trusted me personally and I can tell you that you know I'm extremely confident of the path that we're on and I would urge you to help us put some more wood behind these arrows, so we can take it to the to the finish line (slide 21; 27:18-29:28).