Good morning, and welcome to Needham's 19th Annual Technology and Media Conference. I'm Ryan Koontz. I cover the CommT ech sectors here at Needham that include cloud communications, networking, and optical. With that, I'm really pleased to welcome Applied Optoelectronics, and we've got CFO Stefan Murry. He's gonna lead us through a presentation, followed by some Q&A. And if you'd like to submit any questions via the portal, if you're on the Needham conference chat, you can do that via the portal. We'll try to get to those toward the end of the presentation. With that, take it away, Stefan. Thanks for joining. Fantastic. Thank you, Ryan. I really appreciate the invite and the opportunity to talk a little bit about our story this morning. So, before we get started, I'll just draw everyone's attention to the forward-looking statements and boilerplate language here. I won't read it in the interest of time, but I would caution everybody that you know, to the extent that we make any forward-looking statements on here, there's certain things that you should be aware of, and you can review all those here in the slide as we move forward. Similarly, we have some market data that comes from third-party sources and some from internal estimates, and so, again, you know, just draw your attention to the statement on the risks associated with relying on that data. So with that being said, you know, we'll kind of start the morning off talking a little bit about AOI and what I think makes us special. And, you know, to kind of distill that down into a few points here, number one, you know, I believe that AOI is best conceived of as a leader in advanced optics. And we'll talk about what that means to be a leader and what advanced optics are and where they're being used throughout the presentation. But I think it's worth mentioning that, you know, as a sort of high-level introduction, that AOI has been producing cutting-edge optical devices of various types throughout our history, which stretches back now some 27 years. The founders of the company, myself included, all have academic backgrounds and previous business backgrounds in optical devices and optical communications, engineering, and physics, and that sort of thing. And so, you know, our pedigree as a company and as a team stretches back a long ways and always surrounds cutting-edge optical devices. So we'll talk more about that throughout the presentation, but I wanted to start with that point. We do address some very dynamic markets. The internet data center market, the cable TV market, telecom, and other, you know, specialty optical devices all have some very interesting dynamics. And most importantly, they all rely, to some extent, on ensuring that the technology that companies like AOI provides really helps them keep ahead of the competition. Those markets tend to be very dynamic, and we'll talk about some of the rewards and some of the challenges associated with that as we go through the presentation. We're very fortunate to have a marquee customer list. I would say, you know, most of the large data center operators in the world are, you know, either customers of ours now or have been or are expected to be again. And we also count some of the largest cable TV MSOs and some of the largest telecom operators among our customer list. We're very pleased to have those customers, and we certainly understand the trust that those customers place in our products by virtue of working with us. Part of the reason why we have those customers, of course, in addition to our technology, which I talked about earlier, is our proprietary manufacturing process. And, there's a number of aspects to that, depending on which product we're looking at. But most recently, one of the things that we've seen a lot of customer interest in is our automated manufacturing process for, optical transceivers. And we'll talk about the dynamics of that, how that plays out with our customers, and, and why that's important, later on. But we think it's a very critical part of our, of our story. And then finally, in a diversifying customer base, I mentioned earlier that, that we serve some of the largest operators in the world. In some previous incarnations of our business model, we haven't always directly, you know, sold to those customers, particularly in the cable TV space. So as we've created our own brand name in that space, in the cable TV market, and we've begun to offer our own products under that brand name, that's drawing a more diverse customer base than what we've seen in the past. And we're seeing similar trends, in telecom and other areas where the end users themselves wanna have connection with the companies that make their critical technologies. And so that diversification of the customer base is also something that's an important trend to keep in mind. So before we go into, you know, a lot more detail on, each of these areas, you know, we can sort of highlight, some of the financials and key metrics from the last quarter. Last quarter, Q1 of 2024, we had $40.7 million in revenue, slightly below our guidance range, largely driven by effects of Lunar New Year on our Asian factories and some price reduction that took place in the quarter or took effect in the quarter. On a positive note, I mean, revenue from our 100G products increased 33% year-over-year, and our 400G products more than doubled in that time. And for those of you who have been following our story, you know that, 800G and 1.6T products will be released later this year and early next year, respectively. And we're very excited about the trajectory of those products and, you know, the expected, rapid adoption of that. And I'll talk a little bit more about that, later on in the presentation. In Q1, our non-GAAP gross margin was 18.9%, versus 23.2% in the same quarter last year. Slightly below our guidance range, again, due mainly to, differences in product mix and the price reduction that, that I discussed earlier. On the guidance range, you know, we're projecting a slight uptick in revenue to $41.5 million -$46.5 million. non-GAAP gross margin, a significant improvement there, into the mid-to-upper 20% range, so 25.5%-27.5% on the gross margin. And, that would give us a non-GAAP net loss somewhere between $11.6 and 13.5 million or, you know, $0.29-$0.34 per share. We also are very excited about the trajectory that we're on for the second half of the year. And we'll talk more a little bit about some of the dynamics that are going on there, but suffice it to say that we're expecting, you know, rapid improvement in the second half of the year compared to the first half, and I'm very excited about the progress that we've made in ensuring that that comes to fruition. So, when we talk about advanced optical devices, which was the first bullet point on the introductory slide that I presented earlier. The devices themselves, lasers, photodiodes, transceivers, and various other types of optical parts, pieces, if you will, they can serve a variety of different application areas. So, you know, data center, cable TV, telecom, FTTH. Those areas may seem, from the outside, to be somewhat disconnected. And in some ways they are. The devices that are used in those different applications are not often exactly the same. However, they share a lot of commonality in that, fundamentally, what's driving those markets is the adoption of internet services, cloud services, and, most recently, generative AI as one of those services. But even before that, you know, just video, cloud storage in general, social media, more and more connected devices, all of those trends cause data centers to expand, cause access networks like cable TV, telecom, and FTTH to expand, and they basically increase the demand for the types of advanced optical devices that we manufacture. And so, you know, on the right-hand side, you can see some of the results of that. You know, we've had some sharp upticks in revenue. We've had a few downturns over the years. We're very, very excited about the trend that we're seeing right now. I, I've been with this company since it was founded, as I mentioned earlier, 27 years ago, and I've really never been more excited about the various market opportunities that we have in front of us. Generative AI being largest, but even on the cable TV and even to some extent on the FTTH and telecom side, there's some very exciting trends that bode well for us, and we're very excited about that. Delving just a little bit deeper into these end markets, I mentioned a minute ago that the technology that's used in these end markets is not exactly the same across all of the markets. I did mention also that they do share a commonality in terms of the types of optical devices, fundamentally, that are needed: lasers, photodiodes, et cetera, the types of things that we're very good at making. But the technologies to employ those, I think, are different across the different areas, but they all have strong trends, as I mentioned earlier, that are pushing for wider adoption of advanced optical devices. So in the cable TV space, what we're seeing is cable MSOs, particularly in the United States, and somewhat in other areas in Latin America and Canada, are very interested in upgrading their networks to the newest iteration of the Data Over Cable Service Interface Specification, or DOCSIS, which is basically the protocol that allows high-speed data to be transmitted over a cable network. So even though we talk about this market being cable television, I think it's important for investors to understand that what's driving this market is not television. It's not, you know, analog, linear video. It is the ability to deliver higher speed Internet access to consumers more efficiently, to facilitate things like streaming video, to facilitate things like, you know, communication of generative AI and large data sets and, in some cases, cable television operators are even, dabbling in, in telecom and doing things like, 5G LTE backhaul over their DOCSIS network. So there's a lot of data that's driving the cable television network, and it's not really about video itself anymore. What we know in that market is that some of the largest MSOs are beginning upgrades, you know, just about now, middle part of 2024. Certainly by the end of 2024, we expect several of these large MSOs to be, you know, involved in the early stages of their deployment. AOI's DOCSIS 4.0 compatible amplifier products are just being released really kind of real time now. We had the first couple of production qualified amplifier samples shipped to major customers for evaluation just a few weeks ago. That evaluation is going very well, and we're very excited about, you know, the opportunities that we see both with those products and some follow-on products that we'll be announcing later in the summer and into the fall, and even in the early part of next year. All that is around facilitating the deployment of DOCSIS 4.0 for high-speed internet over cable networks. On the data center side, which I suppose is probably familiar to-... Most of you, the data center market is really being driven by an increase in demand for high-speed compute and high-speed interconnect to service that compute. So, what in turn is driving that compute and interface requirement is really generative AI. So, the rapid and kind of ubiquitous deployment of AI across all sorts of different areas, I think everybody's probably familiar with that from their personal lives or their business lives. But what helps, you know, what makes all of that AI possible really is some extremely fast computing and the ability to rapidly interconnect those compute clusters to facilitate, you know, parallel processing of large data sets. And so you need faster processors, but you also need faster interconnects. And as I mentioned, in the early slides of this presentation, you know, we're currently shipping mostly 100G and 400G products for data centers. Our 800G products have been sampled to a number of customers right now. We're working through getting those products qualified, so that we can see purchases beginning in Q3 of this year for 800G. And then 1.6 terabits is the next iteration. And, you know, we believe that we'll have that product available for market early next year. So very fast product cycle being driven by, you know, I would say, voracious demand by large data center operators for interconnect capacity to facilitate their build-out of AI networks, and that's a very, very exciting trend. Telecommunications and fiber to the home, I won't spend a lot of time talking about them other than just to mention that both of these areas are addressed to some extent by the cable TV MSOs as well. I mentioned a few moments ago that the cable operators are often looking to deploy 5G, LTE services over their DOCSIS networks. And so some of the same products that we make for cable TV, also some products that have, you know, some variations on that theme, are very useful for telecommunications operators, but also useful for our cable MSO customers as well. Fiber to the home is somewhat of a similar story in that, while the cable MSOs continue to deploy and upgrade their legacy hybrid fiber coax networks, in a lot of cases, they're deploying in greenfield situations or, you know, areas where they don't already have network cable. They're deploying fiber all the way to the home. Being able to facilitate a fiber to the home product offering that dovetails well with the cable MSO networks is something that we're very, very good at and very interested in rolling out over the next couple of years. So focusing just on the data center market, you know, I mentioned kind of what we're seeing in terms of our 400G and 800G results. I think it's encouraging that the third-party data that we have, you know, backs that up, right? As an industry as a whole, we're seeing rapid growth of 400G right now. Early stages of 800G in 2024. So for us, again, that will start to layer in in Q3 and Q4 this year, very consistent with the data that you're seeing here. And then we're expecting rapid growth of 800G in 2025 and beyond. This particular study doesn't look at 1.6T. That's relatively early, but we do expect to see some revenue from 1.6T in 2025 and probably an acceleration of growth into the back half of 2025 and into 2026 on that. So, we expect that this market will grow by about 27% a year, and that the total market size here will exceed $10 billion by 2027. This year, 2024, we expect it to be about $7 billion. I mentioned earlier, you know, some of the commonalities between the data center and telecom market in terms of the types of devices and the technology that's needed there. There are some subtle differences in the way that those markets work. I think the data center market is really characterized by rapid product development cycles. I mentioned earlier, this transition from 400G to 800G to 1.6. Those technology cycles are running faster. Those are not necessarily cannibalistic in the sense that 800G coming on doesn't negate the need for 400G, and 1.6 terabits will be used in different areas from 800G. So, they're not exactly cannibalistic, but the fact that we are being called upon by customers to develop these new generations of products at a faster pace is something that's characteristic of the data center market, more so than the telecom market. The data center market has relatively large production volume over a relatively small number of product types, which lends itself to automation, and I'll talk about the automation there in a little more detail later on. But the most important thing to keep in mind is that if you have... Like the telecom market, for example, which is where we and many of our competitors sort of started, because that's where optics was initially used, in communications applications. The telecom market has relatively small volume on any particular product. There's a lot of different products that are needed depending on the network type and the operating environment and what the application is, et cetera. So there's a lot of different products, none of which have individually a very large production volume. And in that case, it's very hard to economically automate the production of those devices because, you know, there's just a lot of product changes, and it's very difficult to design, say, a robotic or automated process to encompass all the different product variety that exists out there. From a planning perspective, it's also harder to plan when you have a lot of different SKUs that you've got to be able to plan for. Both our telecom operators and our cable operators are not particularly good at being able to predict, the exact types of devices that they're gonna need, because it's oftentimes driven by events that happen in the field that are unpredictable. So the data center market, on the other hand, is more predictable in the sense that the operators themselves have, you know, internal plans on what they're gonna do with their networks based on what they're seeing and the service offerings that they intend to roll out. They're able to plan for that a little bit better and give us some better visibility into the forecasting. So again, that also lends itself nicely to automated production because you can, you know, you can plan to have the correct number of robots out there and programmed accordingly and ready to go for production, when it's needed, as long as you have enough time to facilitate any transitions that are out there. And then, the final, you know, bullet point here just sort of gets at the fact that for the data center market, there's a lot of. Because the volumes are high and because the automation is possible, there's a lot of ways to sort of integrate the manufacturing process and make it more efficient. Where again, on the telecom market, with the high diversity of SKUs and the smaller volume in each SKU, it's a little harder to get those economies of scale, and some of the efficiency in manufacturing. So, you know, just something else to keep in mind here. I won't go over too much of the technical and cost details on this. The next couple of slides are really more focused on some of the technology that's used in the data center. You know, you can read those at your leisure, but, I won't spend a lot of time talking about it. I mentioned earlier the 5G mobile backhaul market. Again, I'm not going to spend a lot of time on here. It's a relatively small market for us right now, but I think there's a lot of potential in that, particularly with our MSO customers, because as we move forward with 5G and we start to roll out, you know, more and more, truly ultra-high bandwidth 5G, there's going to be requirements for, different sorts of optical devices, things that can transmit 25, 50, or even 100 gigabits per second over a long distance. And, while we've been producing, say, 100 gig transceivers for years for data center applications, that's a relatively small distance. And the types of devices that are gonna be needed for the 5G networks, tend to need to span larger distances to connect together, elements in a, remotely, located radio access network, okay? And so, you know, again, without getting into too much technical detail there, those types of devices are broadly similar to the devices that we've offered to our cable customers for years. And so, you know, this new emerging application dovetails well with our capabilities. So another key set of differentiators for us, I would say, is all related to our manufacturing and the way that we do the manufacturing and the vertically integrated nature of our manufacturing. And I'll spend a few slides kind of talking about those manufacturing advantages. It's a little different for a technology company, I think, to talk about ourselves as a manufacturing company in a way. But when it comes down to it, there's a lot of aspects to manufacturing that really can give us an advantage when it comes to talking to our customers. And so I think it's important for investors to understand a little bit about some of those advantages and from where they derive. So, the first key advantage that we have is related to our epitaxy or the way that we do our fabrication process. So, I think the most important point to understand here really is that we do in fact have our own in-house semiconductor manufacturing plant here in Sugar Land, which is just outside of Houston, at our headquarters. This process goes back many years. As I mentioned, this was something that we founded the company on this technology, and we've continued to develop and invest in it over the years, to the point where I would say we have probably the most advanced epitaxy facility for optical devices of any company that I'm aware of in the industry. That allows us to do a variety of different devices, whether we're doing distributed feedback-type lasers or electroabsorption modulated lasers, or vertical cavity surface emitting lasers. We have capability to do all of those devices, which means we can offer customers the ability to really choose the device that best fits their needs. And that's not always the case with our competitors, who have themselves sort of locked into one or two of these technologies. If a customer needs something that doesn't fall into that, you know, that technology capability that they have, then it's a little more difficult for them to meet those needs, at least with internally sourced production. We offer that complete capability to customers. I've been talking a little bit throughout the presentation about automation. You know, this is something that AOI has spent about seven years developing, the automated manufacturing for data center transceivers. It's very important now because what we're hearing from our customers is that there's a lot of nervousness among the large data center operators about sourcing components, particularly from China, but even other parts of Asia. During COVID, Chinese supply chains and Asian supply chains in general were more disrupted than U.S.-based supply chains. And with all the geopolitics of these situations out there, you know, there's a lot of ways that their supply chains could get disrupted further in the future, whether it's because of another pandemic or more realistically, because of political concerns or economic situations that may make doing business increasingly difficult in those areas. AOI has really automated the production of these transceivers to the point where, you know, what once took us 1,000 people, for example, in a factory in China, can now be done with 30 or 40 people. And when you can do production in an automated way like that, and decrease the amount of labor content so significantly, then it makes it possible to put that manufacturing in virtually anywhere you want in the world, without incurring significant additional costs because of the labor, the differential in labor rates. That has become a very, very critical selling point for us to our customers over the last, really, you know, several years, but especially in the last 6-9 months. Customers have really resonated with the fact that, you know, you, you need to have an automated production process in order to be able to efficiently manufacture, in a place where labor costs are gonna be relatively high compared to what you would see, in China, for example. Moving on, just a couple of pictures of some of this automation. This is actually a very early stage of our automation. This is a couple of years ago. If you come tour one of our plants now, you'll see that the automated nature of that production is significantly more advanced than this. But it gives you an idea of what we're able to do or what we were able to do, even a couple of years ago. All of this automation that you see, by the way, is developed in-house. We have an extensive team of engineers and technicians that develop new automation technology for us, and we think that gives us a critical and sustainable differentiated advantage over competition in this sense. And so what we've done then with this vertical integration is we sort of replicated that across our various markets, and been able to apply some of the core technologies that we've had across these different product areas. That's really given us the ability to address multiple markets without dramatically expanding our operating expenses, which I'll show you in a couple of slides. A couple of notes on manufacturing and sort of the company in general. I mentioned earlier that we have our headquarters is here, just outside of Houston. That's where we have our wafer fab. We have significant manufacturing operations in Taipei and Ningbo, China. Altogether, the company incorporates just over 2,000 employees currently, although we're expanding that headcount, particularly here in the U.S. to accommodate the new manufacturing that I mentioned earlier. Real quickly on the financials, I wanna leave time for Q&A, so I'm gonna kind of blaze through these pretty quickly because I believe everybody, you know, has a chance to look at our public filings, so all this information obviously is in there. But we did see revenues roughly flat last year. You can see the reduction in revenue, or the segment that contributed to the reduction in revenue was really cable TV. And our data center business actually grew substantially last year. That's on the back of 400G growth, as well as, you know, growth in 100G. And all of that is related to, you know, large data center operators, and it's in some way related to the explosion in interest in generative AI last year. If you look at our quarterly performance, again, Q1 was a light quarter. That tends to be the case. Q1 and Q2 oftentimes are the slowest quarters for us. We saw that a little bit last year. But we are expecting, you know, pretty solid improvement in Q2, and then, you know, a fairly dramatic ramp up in the second half of the year, again, being driven by the adoption of 800G and some of the new customer acquisitions that we've had recently. Again, you know, going through the income statement here, you can see the segment revenue, gross margin, 19% again, operating loss of $17.1 million, and we had cash of about $17.4 million at the end of the year. That I noted on our earnings call last week, that there were several receivables that were outstanding at quarter end, that were actually due, but they weren't collected until early in this quarter. So, that, you know, that cash balance there is significantly improved from what it was at that time. We collected about $13 million in past due receivables in the first two weeks of this quarter, so... Balance sheet here, again, I'll kind of blaze through that, you know, to put a fine point on it. I mean, we think we're in reasonable shape on the balance sheet here, although we, you know, continue to make investments in CapEx and other things, for the expansion that I talked about earlier. I'll kind of skip through that and go straight to Q&A, because I see we're coming up on time here, so- Yeah, definitely. Thanks. That was really great, great summary. Thanks for all that detail. On the data center side, sounds like that's the near-term most important market for you in terms of growing the business here. What are the kind of deliverables that you have ahead of you, and what kind of customer approvals, these sort of things, do you need to achieve your, you know, near to medium-term goals in the data center side? I know you've had some strong business with Microsoft historically. Maybe kind of walk us through what you have to do to execute to secure that revenue. Right. So I think there's a couple of very interesting things that are going on in the data center side, you know, more granular than the introduction that I gave earlier. So you mentioned Microsoft. We did have a couple of contracts that were signed with Microsoft last year, for production of 400G active optical cables. Those, you know, those contracts are expected to provide approximately $300 million in revenue over the next several years, which is the duration of those contracts. And that's very exciting certainly for us. It's a strong mark of approval, I would say, coming from our largest customer, that they expect our technology to be at the forefront of their new network deployments for the next several years. But also very exciting for us is the amount of traction that we've seen on our 800G products. And that's coming not only from Microsoft, but also from other large data center operators. And what's really driving the interest there, again, you know, not to take anything away from our product performance and all of that, because it is very good, but the real benefit that customers are starting to see in working with AOI has to do with our ability to utilize our automated production abilities to produce products here in the United States, here in Texas, at a cost that's not dramatically higher than what it would take to produce the same products in China. So for us, that means maybe 10 or 15% cost differential compared to China. And again, if you don't have a highly automated production process, it's really impossible to produce products in the U.S. cost competitively with China. But when you have an automated process, it makes that feasible, and that's the point that the customers are starting to see about what we offer for them, and that's been where we've seen the most traction over the last six to nine months. Right. So these are growing markets, obviously, in terms of data center investment, but also big share gain opportunities for AOI. Indeed. I mean, I think, you know, we do expect, and we have been taking some share in certain areas, especially in 400G. But as you pointed out, I mean, the growth rate that we're seeing just overall in these markets being driven by generative AI is so substantial that I think, you know, we and a number of our competitors, particularly those who are non-Chinese, I think are gonna probably see several years of really good performance ahead. Got it. A couple questions from the investor base here now. So you feel, in terms of Microsoft in the second half, is that a big contributor to your outlook there, in terms of growth in that account? It is a big contributor, but it's certainly not the only contributor. Microsoft accounted for about 60% of our revenue in Q1. You know, I would expect in the back half of the year that it will still be a significant contributor, but there will be other large operators that we believe will come on board- Mm-hmm ... and will likely somewhat diminish the preponderance of revenue coming from Microsoft. Okay, great. In terms of AI applications, you know, these, these, this infrastructure's been going in for a while here, last couple of years. Is this, is this a, is this a new market entry point for you in AI coming up now? Yeah. I think you really need to look at AI or really all of the data center market in a customer-specific way, okay? It's tempting, I think, for investors to kind of lump all the large data center operators together and say, "Well, you know, they're all offering similar services. You know, they probably all have about the similar, you know, about similar technology needs." And that's really not the case. Different, different operators architect their networks differently. Mm-hmm. The timing of deployments for new technologies among the different operators is different. So I think it's more useful to look at what specific customers are doing relative to their infrastructure. So in that sense, the customers that we're operating with, I think it's fair to say are getting started perhaps a little bit later in upgrading to, say, on 800 gig networks, than maybe other- Right ... data center operators. But that's not taking anything away from their commitment to move in that direction, or indeed, the size and scale of the deployments that they expect to undertake with 800G, and indeed 1.6 terabits next year. Great. Thank you for that. A lot of questions coming in here. We'll try to speed through them. You know, what have you seen as far as the MSO market in terms of their interest in DOCSIS 4? Are you initially focused on DOCSIS 3.1? How are you playing that from a product perspective, and what are you hearing from customers in terms of their needs? Sure. So I mean, again, in the interest of time, I could spend a lot of time talking about that. But, DOCSIS 4, yes, they're very interested. A lot of customer engagement. We have a proprietary technology called Quantum Link, which allows operators to communicate very cost-effectively with elements out in the field. That's driving a lot of the interest that we have in our products. Yeah. Deployments start in the second half of this year, as I mentioned earlier, and we see at least two of the large MSOs, you know, starting new deployments later in this year, and we expect to be parts of, of those. Great. In terms of gross margin, from where you are now to kind of your long-term model for 40%, what are the... What's the bridge? What are the components there to get your margins up in the business? Right. So the 800 gig, and probably 1.6 terabit products in the data center should come in at a higher margin than the current generation of products. Mm-hmm. Layering on more cable TV revenue, cable TV is coming in at a significantly higher margin than the data center. So right now we're in a situation, as you saw in the presentation that I gave, where, you know, last year and the first half of this year, cable TV has been really relatively slow. It's sort of a replacement market right now. You know, while the operators are gearing up to deploy 4.0, they're diminishing their purchases of older 3.1 technology. So we've seen relatively poor performance in cable. Really, you could say over the last year. So that has an outsized impact on our gross margin because of the margin differential between cable and data center. So as cable recovers and as 800G and 1.6T later on, that will help improve the gross margin towards that target model. Cable, I recall you're going direct now as opposed to kind of a resale type model. That is correct. Yeah, that's a very good point to make. Middle of last year, we made the decision to move away from some of our previous relationships with resellers, if you will, or you know, system integrator types and sell directly to the MSOs. That should improve our gross margin by roughly 20% compared to what we had before- Got it ... which is how we get to those really elevated gross margins in cable. In terms of, your cable product or newer cable products anyway, as you look at your second half, I, I assume that's where most of it is. You know, what, what are your assumptions in your guide for the, for the cable business, roughly? We expect to see some pretty strong growth in cable. Cable for us peaked in 2022 at about $35 million in a quarter. Mm-hmm. Mm-hmm. That's kind of where our production capacity was. Now, the 4.0 products are going to come with a price premium to- Mm-hmm ... those, 3.1 products. So for, you know, for the same production capacity in terms of units, we'll probably be able to get 20%-25% more revenue without, you know, adding new equipment or significant new CapEx. So that would mean, you know, we ought to be able to produce $40 million, maybe $45 million, in a quarter in, in cable, without, you know, significant, investments in, in new equipment. And, I'm not saying we'll necessarily get all the way there by the end of the year, but I think we'll make some pretty good progress on it. Great. Quick question here on your 800G product. Question is, is that going into the front end or the back end of the network? I'm sorry, just 800G generally, is that going to the front end or the back end of the network? I'm not really clear on what the distinction between front end and back end. I think those- Okay ... these, the 800G transceivers will primarily be used in interconnecting compute modules within a cluster. So I think that's probably back end by that terminology, but I might be wrong. Yeah. Agree with you. And then, another question: How much revenue in 800G are you thinking next year, and do you think you'll be able to announce another 800G hyperscale customer anytime soon? I do think we'll announce a customer for 800G at some point. It probably will not be until early Q3, I would say, just based on, you know, the qualification cycles that we're going through. We don't want to announce a customer until- Sure ... they've completed all the qualification and what have you. Yeah. So I think there will be an announcement. I'm very optimistic about that. As far as next year, you know, I think just on 800G, you heard on our earnings call, if you were listening to our CEO, threw out a potential number of $500 million-$600 million in 800G for us, in 2025. Now, obviously, a lot has to go right to get to those numbers. But, you know, that dwarfs our current data center revenue, and that's not even counting the 400G and the, you know, legacy 100G business. So there's significant opportunity in 800G ahead of us. Excellent. Well, that's about all the time we have. Stefan, really appreciate your time. Great discussion. Thanks for the overview and the Q&A opportunity. And, yeah, enjoy the rest of the conference. My pleasure. It's been great, and we've got a really good schedule today. So thank you for putting that together, and thank you for having us. You're welcome, Stefan. Take care. Bye-bye.