Sivers Semiconductors AB (publ) (SIVE) Earnings Call Transcript & Summary

Good morning, and welcome, everyone, to the Second Quarter 2024 Webcast for Sivers Semiconductors. My name is Anders Storm, and I am the Group CEO of Sivers. And with me I have...

Lottie Saks, Group CFO.

And we're going to take you through this report. It's going to be a summary, sales overview, financial overview, market and business update, and a summary and a Q&A, as normally is in the end of the webinar. So if we look at the executive summary. We had some great growth this quarter, 14%. Again a stable quarter, over SEK 50 million in sales. I remember back 8 years when I did my first Q2 report, we were about SEK 3 million sales. So it's a significant change in what we are selling today versus what we are selling at that point. And we are coming from very heavy growth last year with 68%. Wireless was the big driver of the growth with 29% growth year-over-year. We are also improving our adjusted EBITDA and our EBITDA, 25% improvement, minus SEK 12.4 million. What is also very strong in this report is that we now see a very healthy pipeline again on the 5G side of things. And especially, 3 projects has been announced: 1 before and 2 new projects today, one with a major global 5G company for CPE. This is maybe one of the most interesting projects we won for the time I've been in the company. It's been a very long RFQ process with hefty competition, but we came out as the winner. Blu Wireless, the track-to-train project as well as a Japanese NRE project with a well known company, but maybe that project is a little bit less big compared to the first one. What is also very rewarding is to see that product sales are up now from -- it should actually say 15%, I think, up to 39% and year-over-year, which is sort of getting us closer to the 80% target in 2026, where we aim to sell more and more product and accelerating the growth due to that. We are also expecting second half to be stronger when it comes to growth than the first half. We have had a little sort of pause in the growth even if we have a good growth this quarter. And if we look at the long-term expected growth, that will be over the coming 3 years an average growth that we are in line with the growth we have had over the last 8-year period, and that is a CAGR of about 40%. If we look at the first half, it was a solid first half with SEK 108 million sales and improvement in adjusted EBITDA by about 30%. We also have a solid cash, above SEK 55 million, and we foresee a less -- much less cash burn also in the second half of the year. Looking a little bit on the growth again, comparing first half year over the last 8 years. We have increased our sales by 16x over this period, which is something we're very proud of. And the average growth over this period has been 42%. And again, I mean, we see this growth being higher in the second half. Always, Q4 is our strongest quarter. But also, this also shows that we are in line with the overall plus 40% CAGR over the long term period. And we have definitely lined up the customers over the 8 years as well as now in this year. We had a very strong sales the first half year when it comes to new customer pipeline. And I will go in a bit more to that, how the pipeline looked like. If we look at the sales then. The Blu Wireless contract, that was a significant event in the sense of we are very committed, of course, to 5G communication but also sustainable communication. And Blu Wireless and Sivers has a very, very interesting track-to-train application that has been rolled out in many places of the world. I'm going to talk more about that. And then the 5G design award from the new customer, I'm not going to talk so much about that today. But just to say that this is probably one of the most significant wins we had so far. And this customer is a well-known customer and we're very, very proud to have won that deal. Another deal we're very proud of is with the Japanese company. We have an NRE project which will also lead into volume in the future here. And it's also a new customer in Japan which is also quite famous by name, at least. So looking a little bit on the track-to-train project. That is sort of a 5G project on the unlicensed spectrum. This is to provide high-speed broadband communication links. And as you know, we have worked with Blu Wireless for this for quite some time. They are now doing an even longer-range solution for us, and we're involved in that and it's a development project. The trains can exceed speeds of 300 kilometers per hour. It's already been tested and rolled out with the current solution. And the contract revenue for this is not very high, but it's a good chunk of revenue that would be completed in 2024 already. If we look at the overall track-to-train systems, to go back to that, I mean, we have seen rollouts. And the most important part is the rollouts that's going to happen now during the fall here with the Caltrain, which is going from San Diego to San Francisco on the new electrified piece. We also had recently a rollout in high speed rail in the northern of Spain, which is sort of start of that. We see many new things. And they've also signed up with new partners where they are rolling out other things in India and other parts as well. And of course, the carbon neutral -- the first world carbon-neutral communication system for trains. So by 2025, evo-rail, who is the supplier, will create that and, of course, reducing CO2 per passenger by train. And you can see here, it's 10x less CO2 used when traveling by train versus air or car. So I will, by that, hand over to Lottie and come back with some more information after that.

Yes. And as Anders mentioned, we had a solid first half year of 2024 when it comes to revenue and with 14% growth in Q2 in isolation. And as you can see from the graphs, the growth is primarily coming from our business unit, Wireless, and from product sales. And we're really happy to see that we're maintaining the balance between the 2 business units with 56% sales in Wireless, 44% in Photonics and with a strengthened split in revenue type, so 39% product, 61% engineering projects. And also, when it comes to geography, we have 28% in North America and 71% in Europe. And I think the main takeaway from this quarter is our ability to maintain the higher level of product sales, which is really important to gradually increasing that and getting to a recurring revenue model. So this is a slide with a lot of numbers, but maybe to repeat a little bit. Looking at the product sales. Growth in total product sales in the quarter is 186%. And looking at Wireless, it's actually not stated correctly here. It's 213%, and in Photonics, 163%. So we're growing product sales in both business units, which is very important. And as you can see, Europe, also strong growth, 53% in total, and growth in Wireless business unit, 27%. And zooming in a little bit on the geographical split, just to recap the differences in our two business units. So in Wireless, we have almost all customers coming from Europe. So 97% of the revenue is European customers. And in Photonics, we are primarily strong in North America at the moment. So looking at the sales in perspective and the trends, how it's evolving. And again, just looking at the step change that we saw in Q4 2022 coming above SEK 50 million per quarter, which is really, really important to uphold. And we have SEK 52 million in this quarter, the record level in Q4 2023. That's of course where we want to be and they come back to that level. So -- but we are still on a good trajectory. And on the right hand side, again pointing out the importance of the mix in sales. To be able to scale on our organization and operations going forward, we want to maintain the level of engineering projects, but then basically grow and leverage through product sales. So SEK 20 million in absolute value is the all time high level and 39% of total sales. So we're really happy about that. But we also want to point out at the same time that we are not in commercial mass production with our customers yet. So there will be ups and downs in the level of product sales. So -- but for sure, we'll see an increased level overall on annual basis. Yes. And then just looking at the profitability as well. So as Anders said, we had a 28% growth year over -- quarter-over-quarter for EBITDA. And actually for the first half, the number is basically the same, so 29% adjusted EBITDA and 31% EBITDA growth for the first half year. So we really see that we are on a continued good trajectory, and we're gradually improving on all measures over time. So we look forward to the future with confidence.

Thank you, Lottie. And from there, I'm going into sort of the market and business update, where we are a little bit. So this is sort of our customer pipeline. And I would say it's better than ever. And we have announced SEK 100 million in new NRE orders -- or plus SEK 100 million this first half year. We announced new design wins with 5G customers that are the best we won so far. And the prospects are really good. We have over 200 prospects. We have qualified leads, 70 of those. We have technical engagement with others where we are sort of testing things. And we have 60 plus ongoing customers. And looking at that value, and as we mentioned before, the total design wins we have when we get into production from '26 to '28, sort of there's a multiple of $100 million possibly in what will come out of that. So -- and if we just look at the first 6 months this year, which has been one of the best ones, you could probably say that we have a possibility to get USD 300 million to USD 500 million in sales based on those 3-year projections that the customer gives us. So NRE in itself is very good, $10 million, $15 million. But however, the possibilities in the product sales is enormous. And here are some examples of customers. And of course some of them are winning, some of them are losing. But in overall, we're seeing a positive trajectory now, as you can see also, with 39% product sales. So when we are getting into the volume phases for real with the customers, that would be a big difference. Just to give you an example, for example, ALL.SPACE, which is our -- one of our most important SATCOM customers. I mean, the content that Sivers provides for 1 terminal is about $10,000. So it's a lot of money per units. ALL.SPACE have sort of started shipping. We have shipped something like SEK 50 million, SEK 60 million. So far we have several ongoing NRE contracts which are moving ahead on different aspects and new chipsets and next generation chipsets and the current chipsets are selling in the first generation stuff. I think that they are a little bit softer maybe on the customer end, what they have achieved so far. But in the long run, we definitely see this as our sort of most exciting and biggest volume customer when it comes to SATCOM so far. Also, we have our Fortune 100 customer. And this is a very, very important customer. We worked with them now for over 5 to 6 years now. The total amount invested is still USD 18 million. They are sort of in the phase where they're testing and making sure that everything works. We shipped in Q1 50,000 chips and they are still under development. But however, the revenue this year is a little bit lower because the NRE has come down on waiting to get to the volume phase where they're doing some testing. So again, this is USD 30 million to USD 100 million per year, which is also a base of those multiples of $100 million that we can get when this gets to fruition. We have not got any information yet when this will get into volume. I don't think it's going to be this year, of course, but we are very positive on the technology that we are providing here. And of course, one of the most interesting markets is our Sivers AI Photonics market right now, where the market exceeds a $2 billion per year opportunity. And we have signed recently with several customers. And Ayar Labs, for example, and the other customer, we signed here early this year. And this is of course a huge opportunity that we will see developing from '25 and '26 and forward, where I think the bigger volumes come from '26 and forward. And those volumes could be really, really high compared to what we've seen so far in everything. And I mean, all in all, putting this together, this is why we have sort of shared this data. We're going from 30% revenue -- product revenue to 80% product revenue in 2026. And we're very bullish on many of those customers coming to fruition during this period. And NRE, we are going to stay the same in that sense. So there is a lot of things happening also on the NRE front as well as with funding from governmental organizations where we're also making very good progress and strides. So going into a little bit of those Photonics things that I mentioned. So Ayar Labs, we -- a year ago now, we signed a deal for one of the steps with them for light sources. And this was a very important project, of course. We have a quite good relationship. We are one of few several sources into their solution. They are, as you know, funded by NVIDIA, Intel and others. We're working now to get this to the next step. It's taken about a year on getting these things to -- and the 16 array working, which is part of this. And we'll demo that soon as well in events that's going to happen this autumn together with Ayar. But then we need to finalize these products and get them into volume, of course. So this is going to support the next generation of generative AI, and we all know how big that market is currently and all the companies that they are working with. The other product we just closed in March, 3 months ago, is with the sort of Ayar Labs type of company but valued even higher and with a wider portfolio than Ayar Labs has today and are working with some very prominent customers as well. This was a $1.3 million project. We are delivering and testing the things as we speak to get to the next phase. And as we mentioned in that press release, we foresee that these first two phases will lead into sort of a medium volume prototype build, which will give us quite a big growth in the start of next year, if successful. And volume manufacturing is around 2026 with millions of chipsets. And of course, with millions of chipsets, if we have a price ranging from $30 to $60 somewhere, that is a huge opportunity for us. And as far as we know, we are the sole supplier to this project. I've been talking a little bit about the flexible go-to-market and asset light manufacturing, and I've got a few questions about it. And we presented in our Capital Markets Day that we're working on this. And at this point, in the end of the year, we deselected 1 of the 3 partners and has 2 partners left that we're talking to and working with. They have both capacity in the range of 20,000 to 40,000 chipsets. One is based in the U.S., one in Taiwan. There's been a scope and cost program down in Q1 and the decision to proceed in March '24. Technology transfer to be able to sort of do all of these things will be happening in the end of '24 as it looks right now. All of these -- both of these partners can actually offer buildup from step 1 to 4. However, we are actually able internally, as you know, to do step 2 and 3 with minor investments. So that is where we're looking at now, how do we do those minor investment versus rest and then sort of getting this to manufacturing in the end. And just to get an understanding, it's hard to understand semiconductors. But comparing standard silicon wafers and semiconductors versus indium phosphide that we have, I mean, the average sales price for a full wafer is about $24,000, while for silicon, they are at $3,000 to $10,000. So at only 6,000 wafers annually, we are at SEK 150 million revenue for our fabs. So the potential in our own fab is also quite large. And the sales here is, of course, with great margins as well since the production cost is 1/8 of sort of the cost of a 12-inch wafer, and of course, lower production equipment cost than in general to build out. So one question that is of course on everybody's minds now when this is my last report. And we have been working on a recruitment and it looks very, very good, and I think it will be announced very soon. In my opinion, the final candidate that is picked is a fantastic candidate, and current expectations is that this person will be able to start before the end of August. And as you know, my last day is around mid-September. So in that sense, we have everything lined up here for a great handover, and I think everybody will be quite happy with the new person who'll take over for Sivers. So again, summarizing everything here. Very good quarter, returning to growth, 14%. Even if we grow 68% last year, we are in a good trajectory. Wireless sort of was driving the growth now and they have been for some time. Adjusted EBITDA is approved and we are moving closer and closer to positive areas here. And as you know, last year the second half, we were positive. And major design awards. The 5G award with a major global company is a huge step forward for Sivers and nothing we've been able to win on that level before. Also getting new traction in that and seeing 5G moving ahead is great. Product sales are up, which is quite important. And of course, we expect to see a lot of things when it comes to much higher growth in the second half as well and expecting to see long-term growth based on what we see right now. So with that, in total, I -- sort of going over to the Q&A. But before I forget about it, I want to thank everybody. It's been a great 10 years in Sivers and 8 years as the CEO now coming to an end. But I'm very bullish about the company, and the company is making great strides in all areas. But it's time for someone else to take over Sivers, and I think we have a very good replacement coming in. So I just want to thank all of you for your support and whatever you supported over the years. But now we're going into the Q&A. So let me get the questions up here and then we'll see.

Any news regarding the contract with the $425,000 from the top tier SATCOM network provider from the 14th of August '23? Yes, that project is going on. It is a little bit delayed. I think we've invoiced about 50% of it. And we are very, very well received from that customer, and that might lead -- that sort of standard solution that we're building for them might lead to some good future sales as well. Any news from the Tier 1 customer of India regarding Intel, WiSig networks? So we are -- WiSig are in a phase where they've sort of built out. They are now at the funding stage of the next phase and they need to get funding to get into production volumes. So that is the Intel stuff. The NXP software is not in any solution today yet. It is actually in a solution that we're looking at for satellite communication things that we're about to sign. But in that sense, it's not up and running. Please update us about future volume, time lines, et cetera, for Thorium Space. So as you know, we signed, was it $3 million second phase for Thorium just here in Q1, which is a very important step. We are working on the production part. And as we said before, the plan is to get that out in 2025. We haven't shared any volumes, but we are sort of -- compared to ALL.SPACE, we expect sort of medium type of volumes rather from them than what ALL.SPACE has been showing. Any news on the antenna-on-display for KREEMO? KREEMO has been quite quiet lately, and I think that's connected to everything around 5G has been sort of a bit more softer, as you know. And we've just stated that we see more and more 5G things happen. So hopefully, we'll see something new coming up there as well when things are changing now. What's happening with the ALL.SPACE customers on next-generation antenna? So ALL.SPACE, as I mentioned already, we are working with them. They are a little bit softer with the current customers, I would say. We are working on -- as we close the $5 million deal on the next-generation stuff, that is of course going ahead and moving ahead. Does the CPE award have anything to do with the previous design win that was lost/canceled? Could it be a similar project like the one that was lost/canceled even if it's another company? It has nothing to do with that company. This company is 10x bigger and more safer company, I would say. And the possibilities is, if they succeed, I would say bigger than that company we had that order with. So that win, as I said, is the most sort of difficult win we had but also a quite large company. How is the test and qualification process of Fortune 100 customers going? Yes, we are -- as I mentioned, we have shipped the chips to them. And I think it's more on their side now how their solution could be finalized on their side. Will Sivers select potential outsourcing partners before a firm order from F100 volume ramp from Ayar Labs, et cetera? Will Sivers select [ also ]? Yes, I think we already -- as I showed you now, we are looking at -- we have selected two partners. And how much we're going to do with each of them is not decided, but we'll go ahead. But the cost for this, this is an outsourcing partner, is not very high. It's more the cost of the qualification in that process. Lately, Intel has spoken a lot about their own silicon photonics solution for data center, which does not look like the previous FPGA solution. Do you know if it's -- Intel has developed a solution by themselves or they are using Ayar Labs solution? So Intel is a large company. They have different pieces. And as you might know, Intel was doing silicon photonics before on sort of a transceiver level. That piece of Intel, which is not connected to the piece of Intel that buying things from Ayar Labs, has developed their own type of silicon photonics product. They are using a solution that is -- where they have the lasers close to the GPU and CPU, which is a different way of doing it in Ayar Labs. And it also locks in if you lose something. So it's not really pluggable optics in the sense that Ayar Labs has. So it's a different way of doing it, but it's in some way a competitor. But we don't know if Intel CPU, GPU side is actually using any of that. Do you know if Intel has their own in-house manufacturing capability for advanced lasers designated in silicon photonics? Yes, as far as we understand, they have a possibility to manufacture those. Why can you never name the companies you're winning agreements with? That's a very good question. It comes down to the customers, that they are very secretive. And for example, in these two latest projects that we announced now in this report, the customers are telling us that they have previous suppliers. They don't want to upset those previous suppliers anyway for the current solutions, and we are replacing those, for example. And therefore, they don't want to mention it. Also, in many other cases, they don't want to tell anyone else what the secret source is and what they're doing, and they want to have sort of the things out to market before anyone else does. So it's a lot of different reasons why customers don't do it. And that is unfortunate for us, because I would say that if we would have been able to mention the companies, it would have been a much easier thing to get everybody to understand how far we have come as a company. Since the Fortune 100 customer put an order for design optimization in January, do you believe that tests that were carried out last year for their sensor did not generate the result that the customer wished for? Or was that their reason for design optimization? I don't know the reason for those. I think that they are on a long-term path to be sort of getting these out. And it's a very important project. It's all the way up to the CEO that is involved in this, and it needs to work perfectly. So basically, they are doing changes as they go. And the timeline, we don't know yet when it is. But hopefully that this is the last round of optimizations, but it could be more. Is the new CEO U.S.-based? I can't be ahead of the press release, but let's see the press release when that comes out. Thank you, Anders, for being a good CEO, and you worked hard. Thank you for that. Yes, thank you, Robert. You are doing really well. Thank you for your leadership. Can you disclose and tell us about your next project employment? So no, not yet. I've been talking to the Chairman of the Board there, and they want something to come out when I'm actually closer to start. It's still a month away. But maybe they want it to come out the day I'm starting or just before. It depends a little bit, but it will soon be really revealed, I think. Why does it take such a long time with the Fortune 100 customer? Please tell us a bit more about that. So the main reason, I would say, is that they are trying to achieve something that never has been achieved before. And this is, in my mind, taking much longer than we expected of course from the beginning when we started this project. And this is connected to what they're trying to achieve. And since that has to be very accurate for different reasons, therefore, it's taking much longer. And we are as disappointed as anybody else that it takes so long time. But still, we are in a very good position to win it. So independently, how long time it takes, we really believe that this is going to be revolutionary, and we are extremely proud to be the supplier of those products currently. Are you working with any F100 companies with data centers? When do you think you will be able to share any news? Yes, we are, and we've been doing that for quite some time. We are also working with other companies around in the ecosystem and hopefully can share more news by the end of this year with certain things. But of course, again, we'll probably not be able to share any information around what's going on, so to speak, and the names. Optical I/O is a huge paradigm shift. What do you think are the greatest obstacle for this shift? Yes. So the greatest obstacles are the integration between the GPUs and CPUs. So they are sort of what we've seen now Intel have done with Ayar Labs. They have their chiplets around their FPGA at this moment. For all other vendors who's going to use Ayar Labs, they need to integrate their silicon with Ayar Labs' silicon and then make that work. And of course, that is a challenge in a sense, but they have come quite far. I mean, Ayar Labs has been around for approximately 10 years now and have a working solution. So I would say that this is sort of where the integration work needs to happen. If we look at the Blackwell platform that came out now, they have NVLink working with copper today. And at some point they're going to do that shift. And of course, that's going to be a huge, huge thing. But I would expect that, that large type of platforms would probably do that in 2026 somewhere. Any news about the LiDAR project customers? Yes, we are still working with them. Especially one of the more prominent ones are in sort of the qualification phase, and automotive qualification is more advanced. So we are negotiating also the sort of MSA contract with them as we speak. Could your laser in the F100 sensor application read several different biometric readings, or do you know if it's focused on just one? No, I mean, we can measure and we have a very broad sort of emitter and detector that can measure. I think it's from 800 nanometer to over 2,000 nanometer. So we can measure a lot of different things, and it could actually be used in very different applications. Do you think that F100 could release the sensor with one initial biometric capability and then continue develop it? If so, support more biometric capabilities? That is an option and depending on where they're going with this. It's hard to say exactly. But it's an option, and I've also speculated that they are actually looking at different type of consumer devices. So we'll see. Yes. So a lot of questions here but -- so it's been hard to select ones, but now we don't have any more questions as I can see here on my screen. It's hard to see sometimes, of course. But do we have any more questions? Otherwise, I think we are handing it in for this time. And again, thank you so much, everybody, for following, and it's been over 130 people on the call. So thank you so much and have a great day, everybody.