Many thanks to Rainer Klute at Agoracom for providing the Q&A transcript from the Annual General Meeting (AGM) a few days ago. I learned a lot!
Here are some snippets:
VApeanut
---------
Adrian Brijbassi: Suresh, we’ll now present you with some questions from shareholders. The first is, do you have enough manufacturing capacity to meet demand?
Suresh Venkatesan: Absolutely. I mean, we spent a better part of this quarter, you know, moving equipment, of course, from super photonics, but also making some additional investments in capital equipment that allows us to seamlessly deliver at wafer scale a million optical engines a year. As the demand grows beyond that, you know, I think the relationships that we have in Penang allows our contract manufacturers to make additional investments for capacity expansion.
So it really places us in a good position with regards to capacity and requirements over the next 12 months and beyond.
.
Adrian Brijbassi: And what steps has the company taken to ensure stability and scalability across packaging, testing, and supply chain partners as those volumes increase?
Suresh Venkatesan: I think, you know, one of the key things, I mean, apart from, you know, all of the political noise around, you know, the geographic issues with China and manufacturing there, you know, one of the things we wanted to do as we transitioned out of Super Photonics into Globetronics was to ally with an OSAT (Outsourced Semiconductor Assembly & Test) that has, you know, decades of experience building semiconductor products.
We intentionally allied with a semiconductor OSAT, because what we’re trying to do is semiconductorize photonics. So what we’ve done is taken a somewhat different approach than traditional OSATs or traditional contract manufacturers.
We went with the fab-in-fab model, where we have our own clean room inside of Globetronics, where we do all the differentiated optics assembly and let Globetronics and the OSAT community handle the electronics assembly that is also required as part of our optical engines. So, you know, allying in that context with people or contract manufacturers that have done this, you know, build billions of units for very large companies, gives us a manufacturing stability, right?
And we’ve been able to seed the manufacturing capability with our capital first, right? And as that ramps up and the business starts growing, then it allows our OSAT partners to make additional capital investments on their own, which is already being done in the context of Globetronics. So I think from a stability perspective, I don’t think we’ve been more stable.
And for the first time, we’ve kind of really made that commitment in terms of semiconductor wafer scale production, which Super Photonics wasn’t going to do for us. And scalability is taken care of by two really well-established contract manufacturers in Penang, in Globetronics and NationGate. And then between the two of them, they have tremendous, not just clean room space, but tremendous capital capability to continue to expand our manufacturing base.
.
Adrian Brijbassi: Okay. We have a question on the CPO module. The CPO module was shown during a 2022 presentation with the recent buzz surrounding NVIDIA’s CPO plans, is there a future offering that POET is still pursuing in that area?
Suresh Venkatesan: Yeah, CPO has been around for a while. We’ve talked about it for almost five years now, and now it’s slowly starting to become real with not just NVIDIA, but also Broadcom and several other customers. And co-package of optics is kind of a generic term, but largely revolves around having an external light source and then your PIC, which in the context of many companies is a Silicon photonics-based solution.
So our presence in the CPO market is to provide the high-value solution into CPO, which is the external light source. And we’ve also been taking our time with our investments as well as our development in that space, because you don’t want to be too ahead of the market either.
But now that we feel that that market is going to be real and it’s going to be large in volume, probably starting production ramp in 2026, but really starting to hit really high numbers in 2027 and beyond, that’s the projections of the industry for CPO. So I think the development of the Smart FAU (Fiber Attach Unit), the development of Blazar solutions, for example, are all going to get ready to intercept in 2026 what we expect to be a relatively dramatic growth trajectory in external light source requirements.
And there are also some other partners that we can work with to generate these modules for us. And that’s another big reason why we did the deal with NationGate. So yeah, POET is getting ready with extremely competitive solutions for the external light source module. And we’re innovating and creating the right solutions that would allow us to be a big player. I was invited at the OFC, as you guys know, to give a talk on multi-wavelength solutions for CPO.
So I think we’re clearly in the mix in terms of solution providers for the space.
.
Adrian Brijbassi: Okay, can you comment on the feedback on Blazar, whether it was received at the OFC meetings or since?
Suresh Venkatesan: Yeah, for sure. I mean, it was exciting. I think it kind of opened a few fairly large players’ eyes into what we’re doing. That was our first prototype demonstration, and we’re now kind of preparing ourselves to make it a real product. And that means not just the engine, but the module, the firmware, the software associated with how the engine works because it’s fundamentally a different laser.
And we’re also trying to build the ecosystem around gain chip providers because we need to make sure that we’re working with people who have the ability to deliver gain chips in high volume for us to have a good solution.
So I think there’s a supply piece to it that we’re working on, and then there are some technology pieces. But I think the interest is unmitigated. I mean, it’s been very, very well received in terms of what we’re doing. I think we just have to put our heads down and convert it to a real product that meets all of the specifications and get it into a module form factor that meets the customer’s requirements in terms of cost and scale and performance.
But I think everything’s pointing in the right direction. Thanks to our ability to be able to pursue multiple things, we’ve been able to get some good people on board and working to convert this leadership position into a leadership portfolio.
.
Adrian Brijbassi: Okay. And can you comment on the relationship with Celestial AI, current status, any delivery of samples?
Suresh Venkatesan: Yeah. I mean, we’ve announced that there’s a production purchase order that we had received and we’re working, like I said in my presentation, to deliver to that PO this year through our partner, NationGate. We’re transferring technology to them as we speak. And so the optical engines will be assembled at Globetronics and then the optical modules will be assembled at NationGate. So yeah, we’re working our way through that.
And then, obviously, we continue to engage with the industry at large on what competitiveness means in the ELS segment and looking at some innovations, like I said, with the Smart FAU, etc., that can be a dramatic differentiator for POET in that space. So it’s not just about the hybrid laser, it’s about how it can be packaged.
Many times in optics we find, even today, that we can make a really good laser and it can be low cost, but then it’s subject to a very, very complex packaging process and a packaging capability and technology that makes it much, much more expensive than it needs to be. So I think our innovations not just address the laser, but also the packaging, and so that the total solution that we deliver is the best for the industry.
So I think we’re making progress there and I’m very pleased with the progress we’ve made. I think a year from now at the OFC, we expect to have it in product form as opposed to in prototype form, which I think would be really, really important for us.
.
Adrian Brijbassi: Now that the facility in Malaysia at Globetronics is up and going, have we hosted any customers there?
Suresh Venkatesan: Several, actually. So obviously, the way I look at it, this journey that we’ve been through, it’s been an idea first, then prove to me that the idea can work, prove to me that this technology can function, prove to me that the product even works. And then you work your way through all of these, prove to me, prove to me, show me. And the final piece is, show me that you can manufacture.
And so we’re at that step now, we’re not talking about technical capability of the technology, we’re talking about can I now manufacture it in really high volume and scale. And so people do want to come and see what we’ve put together. So we’ve hosted multiple visits already and a few more visits planned in July, August, where people do want to come and ensure. Obviously, before they make a supply commitment that includes board and their supply chain, that we actually, I mean, this is real.
So I think it’s all good. And I think the fact that we have this now going, and we kind of waited till about the end of April, because all of the equipment needs to be up and running and brought up. And I think we and Globetronics, we worked with a hectic pace. I mean, it wasn’t that long ago when we decided that we were going to close down Super Photonics and moved all of our equipment out. So I think we’ve moved as fast as we can while advancing our technology and manufacturing.
So we’re happy with that. We’re happy that customers are wanting to come see what we’re doing. Of course, we’ve made the video and so that will be used as collateral as we go through our sales and marketing pitch in the future. So for customers to kind of vet their appetite, if you will, on what it is that we’ve been able to put in place.
.
Adrian Brijbassi: Okay. And we’ll ask about POET Teralight: What is it about POET Teralight that is gaining such interest?
Suresh Venkatesan: Well, I mean, it’s 1.6T for starters, right? So that’s new. Fewer people are working on it, and it’s not yet in high-volume production at any particular module vendors. So a lot of module vendors are looking for solutions for 1.6T and then starting to make their decisions around who they will use and what kind of technology they want to achieve.
So having all of the background with 400, 800, and then putting 1.6T out allows us to kind of get in grounds up into the design cycle. So that’s number one. Number two, we do have the backing of a big player, right? And I think that’s a big deal. So we’re able to get into meetings that we otherwise would have struggled to get into because Mitsubishi is also keen on promoting this laser and its use.
And to POET’s advantage, that laser can only exist in an integrated optical engine form factor. It’s difficult to use that laser in that form using conventional assembly techniques. So it marries that laser to the interposer, which is a good thing for POET, especially since we have someone like Mitsubishi marketing it for us, right? So I think it’s gaining a lot of interest with some players that we wouldn’t otherwise have been able to access, which is really good for us.
And I think we just have to go through. 1.6T is right at the beginning, so I think that’s one big reason why we’re getting a lot of interest there.
. |
