Thanks ITTR

Empire Club of Canada’s Celebrating POET Technologies Inc: Making Money in the Global Photonics Industry.

8:35 Suresh Venkatesan: Thank you. Thank you Antoinette and thank you everybody, including the sponsors, for giving me the opportunity to be here today. What a year it’s been, right? In the first week of January (2020) I was actually in China. We were working through the beginnings of what, today, is our joint venture with SANAN IC: Super Photonics. That was the first week of January. Almost exactly a year ago I just got back from Singapore, and I haven’t left my home since, in a sense. So this is a great opportunity to try to reach out via this medium that we’ve gotten very used to. Oddly enough, I was commenting to Mike (White of IBK Capital), that this has been my first opportunity in a year to actually pick out a tie and wear a dress shirt and jacket.

9:38 SV: I think what I want to do is spend the next thirty minutes or so broadly talking about our innovation, inventions in general, our inventions, and why we believe we’re going to be able to make money in the photonics market. I know the title of this talk was Making Money and it’s somewhat anachronistic, but we’re close to putting products out, and so I think it’s relevant in the context of What is it that we’re doing that would allow us to make a disproportionate amount of money? As you bring more value to the equation that’s what typically happens in this industry.

10:16 SV: That said, let’s go ahead and queue up the presentation. I’ll spend about 30 minutes talking about what we do at POET, and then we’ll open it up to questions, and end the session on that note. What Poet is doing is effectively developing what we call a unifying platform for chip-scale hybrid integrated photonics. As I go through the presentation, every word has meaning. Unifying in the sense that we’re combining the benefits of multiple platforms, that have existed in the photonics world over the past few decades, but we’re unifying them in a means that hasn’t yet been done before. When we talk about chip-scale integration, it means wafer-scale processing. When we talk about hybrid, it means dis-similar die; so we’re not relegated, in a sense, to using a class of die, or a class of chips, or a class of solutions, but we have the ability to hybridize them. And integrated, which means many-in-one, in general. And photonics, which is our primary application. So every word on the slide means something, relative to what it is that POET is doing, and what it is that differentiates POET from potential other solutions that might exist in the market today.

11:56 SV: The premise, fundamentally, that drives success for companies like us, especially in this industry, is invention and innovation. I found this particular statement that Jeff Bezos just made, in a letter to his company’s employees right before he stepped down as CEO, really relevant: “Invention is the root of our success… if you get it right, a few years after a surprising invention, the new thing has become normal. People yawn. And that yawn is the greatest compliment an inventor can achieve.” Which basically says, Look, what we’re doing, especially in the context of disruption, at the point that it is disruptive, can seem: strange, new, hard to adopt. And, over time, it just becomes the way things are done. What we want to try to do at POET is fundamentally change the dynamics of how one thinks about assembling photonics devices, and change it such that it just becomes the norm… people are not talking about it a decade from now, it just the way things are done. That’s the premise of invention. That’s the premise of innovation. So I found that statement to be particularly powerful in the context of what POET is about and what we’re doing.

13:25 SV: So if you were to track inventions over time… and I’ll just use one particular vector: let’s talk about the history of recorded music. It started with phonographs, back in 1890. Then it moved to long-play—LPs—for those of us who actually remember those in the 1950’s. And then we moved to 8-track, or spool tapes, in the 60’s. Then the cassette, or the compact, right? That was one big change. If you really want to talk about the ones that were major—the impact that it had on people, and the one that had an impact from a technology perspective—that transition to compact was important, but probably the most fundamental change was the Walkman in the 1980’s, when Sony put it out. What made that so fundamentally different is it changed the way you enjoyed music; rather than sitting in a place, be it your home or where ever, you now had music on the go. That was fundamental: it was a lifestyle change, right? It was a change in how you did things, how you appreciated music, how you enjoyed it. There was no fundamental innovation other than miniaturization that enabled it, but that change in lifestyle is what made that innovation so dramatic in terms of what happened later. Then came the CD, and eventually you moved to the digital media. And now, finally, we’re in this wave of streaming media. So, in this wave of streaming music, you don’t think about… the other day I was trying to download software, and I had a DVD, but none of my devices had a DVD drive in them… so everything has migrated away from this hard media. You don’t think about it. But at the time you’re making these transitions, these transitions are hard. So when people talk about POET and what we are doing, and where we are in the cycle, and what it takes—and we’ll talk about that stuff in this talk—is we are making a change, making a transition from the status quo, changing decades of how things are done. And so effecting that change is extremely rewarding, but also difficult. That is the journey we’ve been on and we’re happy to note we’re close to putting out products that will finally break this myth that it can’t be done. There’s new ways of doing things that will, we believe, ultimately become the norm in the industry.

16:12 SV: Our vision—which is critically important because it aligns us as a company, it aligns us with our suppliers, and it aligns us with our customers—is to become a global leader (slide says “the” global leader) in chip-scale integrated photonic solutions. And we’re going to do that by deploying what we call our optical interposer, that we’ve discussed before, which provides a seamless ability to integrate electronics and photonics. Integrating photonics is the really important piece, the ability to co-integrate electronics comes for the ride. That seamless ability is what our vision is. It’s why I joined the company: to drive integration in photonics. And we stayed that path; it has been our guiding vision in the company. While the medium of innovation may have changed, since I first joined, the end is still the same. Our mission immediately is to deploy products using our optical interposer solution for the high volume data communications segment of the market.

17:31 SV: So, what is photonics? I think Antoinette did a fantastic job in the beginning, giving an explanation of why photonics is important. Communicating with light is faster and consumes less power, ultimately. And more of the communications medium would like to be using light, if we could get the cost out of the equation, the ability to scale… all of the things that we want to do with our optical interposer technology. So photonics devices create with lasers, detect with photo detectors, and manipulate either in free-space, or with waveguides, in the case of the optical interposer, light. Laser light generated is fundamental to: sensing, computing, data communications, telecommunications… all the biggest trends in computing today. And the ability to move the power generated by the laser, through waveguides, through a detector chain, is critically important. While it might sound simple, it is not, which is why people have been struggling through this process for decades. We believe that we have finally, through a good confluence of semiconductor technology, as well as some innovations that we have done in the way lasers are manufactured and assembled, break through how that entire assembly process works.

19:10 SV: The primary use of photonics is being driven, of course, by the massive explosion of cloud, but also growth in artificial intelligence, as well as the adoption of 5G and edge networks. All of those are basically driving a significant growth in data: data transmission, data consumption, data creation. And as you generate more and more data, the need for optics becomes more. Initially it was basically subterranean optics, or sub-oceanic optics, back in the year 2000. But now it’s optics everywhere. We have fiber optic cables, honestly, in our homes transmitting across rooms. We know the trend is moving to increased use of photonics devices, and that’s why we believe photonics is really important, both today as well as the future.

20:12 SV: Photonics transceivers, which are the primary device that we’re talking about in terms of disruption in the assembly process, are those devices which convert these electrical signals into light signals and back to electricity again. So the way data is communicated using light is: data is first transmitted electrically to the edge of a server, as shown in the picture here (slide 6), there it is converted into light, transmitted through an optical fiber, and then it is decoded back on the other end. And it is that segment of communications, which can be anywhere from 100 meters to several kilometers long in the case of data centers, that has to be done optically because at the frequency and speed at which the data is being transmitted, there is no other mechanism to transmit that outside of optics. So that’s what photonic transceivers are about and that’s what POET’s working on.

21:20 SV: So what is the issue? The issue is that over the past two decades of photonics transceiver manufacturing, the conventional approaches have now proven to be very expensive and very slow. They are typically multi-layers of subassemblies that need to be created and forming the fundamental building block, which is this optical engine that goes into a transceiver. In the case of existing solutions, you start with a laser chip: you first mount it, create a laser subassembly, then that laser subassembly becomes a transmit subassembly, similarly you do that on the receive side, create another subassembly, there’s a housing sub assembly… it is insane how complicated that process is. But, nevertheless, that’s the process that’s been used for the past twenty years and so there’s always a resistance to change, and there’s also a burden of proof that an alternate way of doing things can in fact be quite competitive. But this process is not scalable. It’s expensive, just because of the number of steps involved, no matter how many people you throw at it, and no matter how many garage-stop operations you set up to create all the sub assemblies, at the end of the day you consume time, you consume capital resources, they’re expensive to build and far more importantly it’s difficult to scale. So when we’re talking about really high volume applications, being able to cross that chasm from building a million units a year to building a hundred-million units a year, that’s a degree of scale increase that you can’t go linearly, and there needs to be a supra-linear way of addressing it, which is where integration comes into play.

23:14 SV: This is not new—I mean, people have talked about integration since the early 2000’s. Yet it has not yet been implemented, despite many companies working on it for a couple of decades. I, personally, have been working in this field, since the early 2000’s, with silicon photonics. I’ve known of some of the issues that are there. And what we have done at POET is to try to architect perhaps a different way of looking at the problem, re-thinking in a sense how assembly ought to be done, and coming up with a solution that is not only achievable, but also scalable. And scalable not only in terms of performance, as well as scalable in terms of cost.

23:55 SV: So why does integration matter? I covered some of that already, and especially in the world of photonics, integration matters a lot, and especially integration of the assembly or packaging aspects of the photonics device matters a lot, because much of the costs of photonics is in the assembly process: assembling the lasers, insuring that the laser light is coupled correctly, that all the light from the laser is usable and not wasted along the way, and insuring the laser maintains its fidelity as it’s being packaged, that it doesn’t deviate from its initial specifications as it’s being packaged. These are extremely esoteric components… they are finicky, you know even among the class of semiconductor devices, optical semiconductor devices, these devices are hard because they use non-linear physics in the way light is being generated and emitted. So dealing with these has been difficult, which means the packaging has been extremely difficult, as esoteric, and expensive, and what we’re trying to do is to de-mystify that process and apply more conventional semiconductor technology to that process. There’s obviously skepticism, which is why we’re going through these multiple prototypes and validation processes, to get past that and to get to a point where customers actually say, Aha, this can be done, and we’re moving down the path to integration.

25:43SV: So our approach to integration is not dissimilar from what others have done in like-spaces. In this particular example it’s an iPhone—or any smart phone for that matter—and here the integration was in functionality… so you can integrate a camera, a GPS, and so on and so forth. For us it’s the integration of components. So that’s one aspect of what we’re doing: integrating hybrid components together into a singular chip. The other aspect of what we’re doing is to create a platform. A platform is critically important when you’re trying to access multiple applications because you really don’t want to do a grounds-up design each time, to address a new product portfolio. There’s always going to be some extent of re-design required, but if you do it in a way where that re-design is minimized, one has the ability to proliferate known solutions across multiple segments, either within a similar vertical, or across multiple verticals. So we have taken a fair bit of time, upfront, to develop within the context of a platform to enable that kind of re-use to occur.
27:08 SV: So, in short, POET is doing for photonics what semiconductors has done for electronics. We’re achieving lower costs, higher performance, through device integration and wafer-scale fabrication. It seems—sounds—so simple, but applying it to photonics has not been. We believe we are at the cusp of this breakthrough, where we can actually put products in customers’ hands, that have been built in a very, very different way than the way people have built these products for the past thirty years. It is meaningful and it’s significant.

27:54 SV: The optical interposer platform, like I talked about earlier, is a unifying platform. There have been platforms for integration in the world of photonics that have existed, particularly for passive devices. Or interposers have existed. Or microoptic assemblies have existed. But what we’ve done at POET is to figure out a way to unify the capabilities of these platforms to what we now call the Optical Interposer Platform. So we have the ability to address many market applications, where as silicon photonics is typically relegated to a few. We do believe we have the ability to address a very wide range of applications with our technology in the context of an overall unifying, integrated platform.

28:55 SV: I think the key is (now moving from) platform to products; a product platform is not a product, so what we’ve done so far is develop the platform. We are now in the process of converting that platform into actual products. We’ve made multiple announcements with tape-outs, and imminently in the process of being able to place these products in the hands of customers for evaluation and feedback and so on. It is a collection of common elements. We do expect that most of our products will have a set of common elements, and then there will be new products that will have, potentially, different elements, but with a common backbone. We expect, therefor, to… you will see that in our roadmap as well, we go from 0 to X number of products almost simultaneously, only because we’ve developed it in the context of a platform, which then allows us to proliferate that platform across multiple segments.

[30:05] So what are we doing? I mean we are rethinking conventional photonics assembly, I think. And our standard assembly processes gives you a sense of how much equipment, how many steps are needed. And what we tend to do is to basically move to a scheme where due to wafer level assembly and the integration of all of the passive components into that wafer which we call the OI, we reduce the number of components. We reduce the amount of capital. We reduce the amount of time. I have been to some of these operations, in China particularly, because that’s where much of this gets done these days. And its basically wall to wall equipment and people. Each person is sitting and handling this assembly process. And its extremely tedious, time consuming and expensive.

Now of course, over 20 years, this has become an operation. And people just know how to do it. You can almost turnkey one of these operations from start to finish, to place a whole factory like this in a year. But it doesn’t scale beyond a certain point. Because there are sufficient fixed costs. It just doesn’t go away. You know five years ago, for example, people talked about 100Gbits/sec. It needing to be at $100.  $1 per Gbit/sec was like a mantra that was being used. And you know several people went about doing 100Gbits/sec, and today the prevailing price in the market is still about $120 to $150 per module depending on who you talk to. So you know it never really got to that [$100 per] 100GBits/sec point. And we know that with our technology we can break that barrier. We can actually get to quite a bit below that $100 barrier. So that’s meaningful. That’s because we eliminate all of the stuff that you’re seeing there on your screen [transceiver manufacturing]. And we migrate to work at wafer scale capability with fewer tools, fewer capex… but it is a very different way of doing things.

And I think one of the challenges we face is… it’s a question I get asked all the time, I’ve already placed all of this infrastructure I already have all of these tools. What am I going to do with it? So that kinds of harkens back to that music analogy At the point of disruption it is sometimes painful. Especially the folks that are kind of used to doing things a certain way. But over time it becomes the new normal. And that’s what we hope and expect to develop as we proliferate our technology in the market.

[32:56] As I’ve discussed this fundamental building block is called this transceiver chip. And so we are rethinking the transceiver itself. We’re applying this interposer concept, applying the wafer scale capabilities. We’re effectively able to devolve these many sub-assemblies and many components that exist into a single chip implementation. A single chip implementation that is capable of up to 50Gbits per second, per channel, for 400G applications. The same technology is available for 100, for 200… it’s a very scalable platform. You know first there’s difficulty, I mean we’re talking cutting edge.. not even cutting edge, bleeding edge in terms of what we are trying to do. But we believe that the simplicity of the capability and of the integration is what ultimately is what is going to make this a very successful platform. We use simpler bill of materials, reduced steps, using standard assembly technologies, and something that is scalable. Scalable in terms of capital, in the sense its super linear when it comes to the number of devices for a certain amount of capex that has been invested. Which is what make this attractive, relative to the more straight forward linear kind of approach. Which is “oh you need 1000 more. Let me put another station with another person and another tool”. And we need to break through that mold if this is going to be an attractive technology for high volume.

[34:37] So as a consequence of developing this capability, we have designed, and its in the path to getting productized today, what we believe Is the world’s smallest, lowest cost 100G optical engine. We can put the lasers, photodetectors, monitor photodetectors, multiplexers, demultiplexers, power taps… all within the 9mm by 6mm. I mean you could probably measure 9 by 6 on your fingers [pinching fingers together]. I mean it is pretty small. But within the scale of what we’re talking about you can fit four of these within the space that today is occupied by one right, which is huge. There are some customers that we are talking to that want to leverage the fact that we can make it quite this small. And leverage that form factor by doing some things innovative in the module that haven’t been conceived about before. And that’s what this innovation is in a sense that for different people it could mean different things. For some people it could mean “Oh my God I can get the sub dollar… hundred dollars for a 100Gbit link.” For some people it is “Wow. This form factor is so small I can do something that I couldn’t do before.” We are exploring that with multiple parties as we start proliferating our capabilities across kind of the value chain of customers. And we’re seeing that resonate, which is, you know, it could be different things to different people just because the dimensions in which we bring value for the customers.

[36:27] So what are the benefits of the IO. You know the top line is cost. We think that the way we go about doing things, when scaled to full volume and running the way normal semiconductor assembly runs, which is very high yields, we expect these module costs to be substantially less. I mean these are not incremental changes in pricing, which is what people normally see, which is an attritional rate of three to five percent a year, we’re talking about a fairly large one time reduction. Once you taste that fruit, it’s kind of hard to give it up. We need to get to that point because I think until such time as that happens, people are going to be very happy doing things the conventional way. It’s lower capex investment for module assembly and test; I’ve discussed that. It’s chip scale packaging which is wafer level integration. And it’s a hybrid planar technology which allows us to use dissimilar components. And for different applications that could be of value. I think today people don’t know what that means because they haven’t had the luxury of being able to do it that way. And I think that there’s some education that says “hey, for this particular application perhaps you can use this device instead of this device,” because as far as we’re concerned from an interposer perspective its just a don’t care. So lets pick the best devices for the applications and that’s that flexibility that we think is going to be critically important going forward.

[38:13] So over the course of this quarter, and I think its one of the reasons why we’ve been boisterous, I think we’ve come out in a sense and have discussed this openly is we’ve generated a fair number of industry firsts. I mean which is saying something. We have been at this for about three years, I think we conjectured about the interposer in 2017. We announced the first platform in 2018, or the concepts. So we’ve been at it for about three years, ok. And in three years we’ve generated data better that people have in 20 years. After billions of dollars of investment. Not that I couldn’t deal with a billion dollars of investment, but the point is when you think of doing things differently from the ground up you can make disproportionate progress, which is how innovation works. That’s how invention works in general. And that’s what we’ve done. So we’ve got the industry’s smallest transmit receive optical engine. We’ve got the best coupling efficiency and I’ve discussed this in previous meetings as well. Our losses through our system… to usable power is the highest you can get in any photonics assembly, even done in the conventional way. We’ve also demonstrated the first directly modulated laser that has been flip chipped and assembled at wafer scale. And that is important because a) nobody has done it, and b) doing that opens up the door for 100 and 200G applications that use directly modulated lasers. At 400G and beyond we’re using continuous wave lasers, but at 100 and 200G which is the biggest volume of market today, requires DMLs. It’s really critically important that we’ve been able to achieve that milestone, which is why that was a big announcement. And it had the right effect. It resonated the most as well with the investors, with the group, the customers.. I mean it was a meaningful achievement for us to be able to get there. And of course that comes along with the full integration of the multiplexer, the demultiplexer, that we’ve talked about for multiple years now. It’s taken us a fair bit of time to really get that level of performance that Is needed for these extremely stringent requirements. And finally we do believe it’s the only one of its kind today, a fully hybrid photonics assembly platform. We believe, at least intellectually, you can project out and say “Hey you know all of these components work you can project intellectually out and say this can be done. It can achieve these levels of cost. And yes there’s the burden of execution is on us to be able to do it. But we believe that we’ve gone through all of the key requirements

[41:30] People ask me all the time, “Why are you in this never ending prototype loop?” So I thought it would just be important to kind of educate and kind of bring back the basics in a sense. What does a design process look like? The design process starts with kind of understanding what the market requirements are, defining the technology and then innovating. So the entire innovation piece went about during the 2018 time frame for us as we re-architected for us what we believe is integration in the space. And then we’ve been in this prototype world for a while. But even in the physical prototyping concept, there is always a proof of concept prototype that basically says, “Hey look I’ve validated every individual block.” Fundamentally the proof of concept has been ticked. We did that about a year ago. And you know now we’re in this kind of functional prototype demonstration phase where we’re able to articulate it’s the worlds best this and industries best this. These are functional… meeting specs. But not completely meeting all of the requirements, there functional. And now the next step is alpha which are what in this particular slide says “golden samples”, dry run prototypes. These are basically alpha samples and that’s what we’re in the process of doing. We taped the products out late last year. And we’re running through the fab over the course of the next couple of months. Those are the ones that go in the hands of customers. So far our engagement with customers are databased. And we get to the point where people actually have these golden samples and prototypes. Which then once reviewed and approved go into production. And so that’s the final phase if you will of this prototyping we’re in and I know several investors have asked me over time, “What is with these prototypes and samples.” And we tend to intermix these terms, and for us its easy, we’re in the industry. We know. We’ve kind of gone through the process many times. But I’m trying to de-mystify this in a sense as to where we are. We are in the final stages of culminating, if you will, this entire ideation, innovation process ito something that’s going to generate value for customers. And it’s going to generate value for our shareholders. And it’s going to generate value for the company.

[44:11] So our product rollout strategy remains unchanged. You know it’s a scalable optical engine platform for multiple frequencies. So these are the modulated laser based products which are the 100G and 200G. And then we have the CW laser-based products which are 400G starting with the receive engines and eventually moving to the modulated engines. We expect over the course of the next month to make some announcements around partnerships around 400G. You know I think most of you know we make the optical engines, we make the lasers, but we are partnering with other parties on the modulator. And so because it’s a platform technology that is hybrid, we have the flexibility of multiple partners for the modulator and we expect to be working through that process over this next quarter.

[45:10] So our roadmap remains unchanged. I think we are in this kind of alpha sample phase. We expect to get into production by the end of the year on the 100/200G. We expect to be in production by the end of the year on the 400G light engine as well as the receiver for 400. So this is one of the advantages of the platform, is that we do expect readiness in terms of being able to ramp it in manufacturing in that period of time on multiple products. Of course, we can’t do that with the skeleton crew that we’re in today. We’re happy to announce that we’ve closed this round of financing with support from some of you.. all on this phone call. And you know it’s important because we need to scale this company. I mean we can’t be in this 20 person R&D mode and expect to generate the kind of revenue that we believe we can. It requires the company to scale. We’ve got to mature. We’ve got to be a player in the industry commensurate with the kind of technology that we’re talking about.

[46:25] So we’ve talked about the JV, with Sanan. The JV with Sanan provides manufacturability and scale.  And finally the opportunity. We’re working on data transceivers for the data center market today. We will be working on 5G. We’ve already announced a co-packaged optics solution especially in the artificial intelligence space for optical computing. Each of these id a fairly sizable opportunity for us. Because what we are doing is fundamentally revising the way photonics packaging ought to be done. Its like saying, “You’ve got a fundamental packaging technology well where can you apply it to?” Well everywhere. But where are the ones we think we can add more value now and in the future? These are the verticals [transceivers for datacom, 5g, co-packaged, optical computing] we believe we can. We believe we can make meaningful penetrations in these vertical segments. But of course our focus initially is transceivers for data communications.

[47:33] So in summary, I think Poet… we’ve successfully believed transitioned our company from technology development to products. We’re now imminently in this final phase of prototyping, getting golden samples built. You know, engaging customers with golden samples that meet their requirements. What we have seen from our functional prototypes are exceeding our expectations in many cases. So we’re extremely pleased to see that. We’ve established our manufacturing arm which is Superphotonics. And we believe this company has the opportunity to build a significant revenue business. Which is what many of us want this company to do. You know several of you have been extremely supportive in our mission to get there. I am extremely pleased with the progress we’ve made over this year. We hope you see it as well as we communicate this message to you. And I look forward to your continued support as well.

Questions:

What initial activities or milestones have been achieved to date with Sanan?

I think there’s been a fair bit of activity in terms of setting up the JV. I think the infrastructure in terms of the building, the space, that’s all been done. We’ve hired the management team. The board has been established. Now we’re kind of in the process of acquiring capital. As you know the JV is set up so that the capex is procured from Sanan. And so we’re now in the process of placing POs and buying capital. We expect the JV to be functional, in terms of being able to do the assembly, from about the second quarter. So til that time we’re still working with several subcontractors to do the work we need to do. We expect to continue that process through such time that we generate these alpha samples It’s critically important. We do that so we don’t want any disruption there, but we would expect that transition to occur around the 2nd quarter of this year.

Can you frame what Poet believes the ultimate revenue potential is of the JV?

For the JV right now is set up primarily for 100, 200 and 400G data communication, so obviously the revenue potential there is limited to that market vertical because that’s what the JV is started to do. Of course we have the ability to expand the JV into multiple verticals and increase it’s revenue potential. But for the market segment of 100, 200 and 400G data communications, the revenue potential for us collectively is about a quarter of a billion dollars.

Can you provide insight into the scope of discussions with prospective customers?

You know we are hampered because we can’t travel and visit them, we think it’s far more meaningful and powerful to be able to do that. But we’ve had engagements with several customers since the beginning of this year because we’ve generated the data package we believe is important to have these engagements. When you are early in your development you kind of limit your visibility to a few. As you kind of progress through your development the burden of data is more and so I think now we’ve generated the data back, it’s necessary to engage meaningfully. And we’ve been pleased with the take up. I think a couple of years ago we made this announcement with a large module company Accelink. Obviously we’ve had some delays on our end, getting solutions out since that announcement but we’ve re-engaged and I think those engagements are obviously positive. We feel we’re in a good spot. I don’t think at this point in time we’re going to be demand limited honestly. I think we’re going to be.. out limitation is to overcome, the ability to get samples into the hands of customers which is a process. I mean there’s no major science that needs to be re-invented at this point in time, but we have to go through the process. We’re taped out. The fabs are doing what the fabs need to do. And we go through the process. But I don’t believe we’re going to be demand constrained. I don’t think we’re going to be at a point where we are searching for customers to adopt our technology. That’s going to work itself out thankfully.