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

Good day to everybody, and welcome to the Second Quarter 2026 earnings webcast with Sivers Semiconductors. My name is Vickram Vathulya, I'm the CEO. And I'm joined by Heine Thorsgaard, who's our CFO. The agenda is as follows: I start with an executive summary, then I will hand it over to Heine for financial results in a deeper dive. And I'll come back for a business update, followed by key takeaways, and then we'll open up the floor for a Q&A where people can submit written questions. Our opportunity pipeline has expanded to $1.2 billion as of July 2026. Our second quarter revenues came in at SEK 53.8 million or SEK 55.6 million at constant FX, and our adjusted EBITDA came in at negative SEK 35.5 million. Softer H1 2026 financials, as indicated. Our product revenue was up 18% year-on-year at constant FX. In addition to the government-related delays that we talked about last time, we have made a conscious resource shift by prioritizing product ramps versus NRE, and our revenue inflection is expected in Q4 of 2026. Moving on to highlights. Exceptional opportunity pipeline momentum continues. The July 2026 pipeline at $1.2 billion is a 268% increase from the end of 2025. We've also identified $4 billion of new serviceable addressable market for our semiconductor optical amplifiers [ who are ] used in optical circuit switches inside AI data centers. We are executing well on multiple product ramps for 2026-2027. Production [ builds ] are ongoing for Tachyon Networks. We have now received production orders from ALL.SPACE and initial program orders from SemiNex. Production orders are imminent for our LiDAR customer, and solid prospects are emerging for revenue from pluggables in 2027. We have made an important strategic shift for our Photonics business from a fab light to a hybrid manufacturing model that I'll talk about later. And this is to bring on board fresh capacity and global supply resiliency for indium phosphide lasers, and our Glasgow site expansion is underway. With that, let me hand it over to Heine Thorsgaard for an in-depth discussion of our financial results.

Thanks, Vickram. In the next few slides, I'll take you through the financial performance and key drivers for the quarter, touch on the significant strengthening of our balance sheet and financial position and explain how we are positioning the company for the next phase of growth and scaling. The key highlight for Q2 is straightforward. We are deliberately transitioning resources from NRE projects towards upcoming product ramps. While this affects near-term revenue, execution remains strong, product revenues are growing and our pipeline development continues to accelerate. Turning to our Q2 financials. Before looking at the total revenue number, the most important development in the quarter is that product revenue increased by 13% year-over-year or by 18%, adjusted for FX changes. This reflects the transition we've discussed over the past quarters, where development activities are moving into the final stages before customer production ramps. Total revenue for the quarter was SEK 53.8 million compared to SEK 61.4 million last year, a decline of 12%. Adjusted for FX changes, the decline was approximately 10%. The primary reason is the deliberate reduction in NRE activities as customer programs progress towards production. This is an intentional shift in resource allocation rather than a reduction in customer engagement or demand. In addition, currency movements customer timing effects and delays caused by the U.S. federal defense budget's postponement early in the year continued to impact the quarter. On top of this, Q2 costs are significantly impacted by a noncash accounting effect linked to our share price. During the quarter, the share price rose from SEK 10.71 to SEK 63.15. And as a result, we recognized a SEK 42.9 million social security expense related to employee share-based incentive programs. Under Swedish regulation, we must accrue employer social security charges and the calculated value of employee equity awards and remeasure that liability against the share price at every [ rebelling ] date. So movements in the share price can swing reported operating expenses meaningfully from quarter to quarter. Let me be clear on two points. First, this expense has no impact on operating cash flow in the quarter and does not impact the underlying performance of the business. Second, when share options are eventually exercised and actual social security payments fall due, we hold treasury shares to be sold in connection with those exercises, and the proceeds will be used to offset the social security payments. So effectively, the social security payments do not end up impacting our cash position. With that accounting effect set aside, let me turn to the underlying earnings picture. Adjusted EBITDA was negative SEK 35.5 million compared to negative SEK 20.9 million last year. In addition to revenue timing effects, this result reflects investments in future growth, including product ramp preparations, sales expansion, operational readiness and our continued preparations for a potential U.S. dual listing. As previously communicated, we are evaluating a potential U.S. listing as part of our long-term capital markets positioning and ambition to broaden our investor base. Our readiness work ensuring that we have the required financial reporting capabilities and governance framework in place is progressing according to plan. One of the requirements for U.S. listing for Sivers is 2 years of financial statements audited to PCAOB standards. In early 2027, this is expected to be behind us, and all necessary preparation work is expected to be completed. From that point, a potential listing becomes a question of decision and timing rather than readiness. Overall, Q2 should be viewed as a strong transition quarter where product revenues are growing, customer programs are moving closer to production, and we continue to build the foundation for future scale. Now let me turn to the balance sheet. During the quarter and shortly after quarter end, we completed a series of transactions that fundamentally strengthened Sivers' financial position. In total, we raised SEK 825 million of gross equity capital and converted the USD 12 million bootstrap convertible loan into equity. Note, however, that the completion of the larger June raise and the loan conversion happened in July and therefore, not shown in the balance sheet at the end of Q2. Together, these events materially strengthened the balance sheet during July. More importantly, they provide the financial strength and flexibility to support upcoming production ramps and enable us to invest proactively in additional manufacturing capacity. The key takeaway is that we no longer need to focus on funding our transformation. Our focus is on disciplined execution, capacity expansion and on converting our commercial opportunities into long-term shareholder value. We are increasing investments in our own manufacturing capability because qualified [ merging ] indium phosphide laser capacity is scarce and increasingly supply-constrained. By owning critical process capability, we gain greater control over capacity lead times, quality and supply chain resistance. This is particularly important in a supply-constrained environment as our customer programs move towards large-scale production. Let me then turn to how we think about the company's development and path to scale. As communicated previously, we're in the middle of a structural transition from an engineering- and development-led business to a product-led company. Q2 is a direct reflection of that transition. Product revenue grew by 13.4%, while NRE activities were intentionally scaled down as customer programs move closer to production. As highlighted before, we view this transition in three phases. The first phase was validating our technology, building customer relationships and establishing a strong and growing pipeline. The second phase, where we are today, is focused on moving programs into production, supporting initial volume shipments and increasing the share of product revenue. We are seeing tangible progress across both our Wireless and Photonics businesses, and we continue to see strong momentum in our opportunity pipeline supporting future customer ramps across all key end markets. As we execute through this phase, revenue becomes less lumpy, visibility improves and margins and earnings strengthen over time. The final phase of the transformation is where product revenues, gross margins and operating leverage become increasingly visible in the financial profile. The key point is that value creation does not happen in a single event. It happens through a series of customer production ramps each improving visibility, mix and ultimately, profitability. As we move into 2027, we expect these ramp dynamics to become increasingly visible. Our opportunity pipeline is record high, providing support for the next phase of growth. So to summarize, Q2 reflects a deliberate transition, while underlying execution remains strong, the balance sheet is significantly stronger, and we continue to position the company for a transformation of 2027. With that, I'll hand it back to Vickram for the business update. Vickram?

Thanks, Heine. Let me take the audience through our business update. First and foremost, we are riding a historic semiconductor upcycle. Record industry demand led by AI and datacom infrastructure support Sivers' focus areas. June chip sales came in at $134.5 billion, and 2026 global sales for semiconductors is projected at $1.5 trillion, which is a 90% increase from 2025. And it is projected to sustain that momentum going into 2027. We play in specialized high-value segments, indium phosphide photonics for AI data centers and beam forming wireless for SATCOM and defense. The same AI and datacom infrastructure CapEx that's driving the broad market directly expands Sivers' addressable opportunity. Our technology supports three super cycles: AI factories with our photonic solutions and SATCOM and modern warfare with our wireless solutions. These secular trends are ones that override traditional semiconductor industry cycles, providing longevity and sustainability of our strategies. Let me remind our audience of the Sivers opportunity pipeline. This is a lead indicator for future revenue growth potential. Leads and raw opportunities become qualified after we've had several discussions with our customers. We have gained conviction in the strength of that product and the fit of our technology within their product. And the customer has given us line of sight and early forecasts of their product rollouts, at which point this becomes a quantified number as well. We tie the time line to our strategic planning horizon when we dollarize the pipeline and share it with our audience. Our strategic horizon is a 5-year horizon. So at the end of 2025, that horizon is 2026 through 2030, which is the time frame over which we show our opportunity pipeline, which is nonbinding revenue potential. As we get to the end of 2026, this will roll forward and become a 2027 through 2031 pipeline. We're also consciously migrating from true custom products to standard products to help us accelerate the conversion cycle from a qualified opportunity through a design in and design win leading up to the production phase, which will be a multiyear production phase. With that said, our opportunity pipeline has grown to $1.2 billion. That's a 268% increase from the end of 2025, which is a monster increase. Just from May 2026 to July 2026, the pipeline has increased by 52%. Strong momentum in both Wireless and Photonics, and as Photonics opportunities continue to come into the pipeline, we soon estimate that the Photonics contribution to the overall pipeline will even be better and higher than the Wireless pipeline. Let's talk about some of these product ramps coming online. Fixed wireless access products are on track. Tachyon Networks is already inside a multiyear production cycle with us. We are executing to their initial production order, and their customer pipeline is expanding. Our Tier 1 telco vendor is an advanced systems integration in trial phase, with product release slated for the end of 2026 and end customer trials to follow thereafter. This will also be a Sivers product revenue contributor 2027 onwards. We have now received the ALL.SPACE production order for 2027, and this is a serious commercial validation of Sivers' wireless beam formers and the ALL.SPACE Hydra 4 terminals. This is what I call a best-in-class value proposition and also signals the start of a multiyear production cycle. You can see that this is going to become a repetitive theme, where initial production orders are layering in to start multiyear production cycles. ALL.SPACE has strong traction and pipeline growth with the U.S. Defense, Army and Navy. And as I've said before, when the U.S. Defense starts deploying these solutions, the NATO nations follow. So currently, ALL.SPACE is in trials with the Canadian Navy. And the Hydra 4 terminals support multiple satellite constellations such as Viasat, Telesat, Amazon, Kuiper and many more to follow. This increases the attractiveness for increasing deployments in the field across geographies. Our automotive customer ramp begins Q4, and initial production orders are imminently expected for Q4 2026 and 2027 from our strategic customer. This imminent order again signals the start of a multiyear production cycle, and this will be a strong revenue contributor for future years. Our organizational partnership between the two companies continues to strengthen at the topmost executive levels to maximize collaboration potential across use cases, applications and markets. Within AI data centers, Sivers' CW laser momentum continues to build across product categories and use cases for both our discrete lasers and laser arrays. I'm trying to simplify the picture here for our audience. On the leftmost graphic, we see a traditional pluggable transceiver, which we have opened our aperture to support since a couple of quarters back, where we were focused almost exclusively on co-packaged optics. For traditional pluggable transceivers, our production-ready 70-milliwatt and 100-milliwatt laser singles and arrays are the right fit, and we are seeing tremendous traction, which I will outline on the next slide. On the rightmost part of the graphic is co-packaged optics, where all the photonics gets co-packaged alongside the GPU or the switch ASIC, leaving the external laser or light source outside, either as a pluggable or as an onboard margin. We continue to stay engaged with partners and customers in co-packaged optics. And here, the needs vary and are highly architecture dependent. The need for the lasers varies anywhere from 200- to 400-milliwatt lasers with and without semiconductor optical amplifier combinations. So we continue to monitor the architectural evolution and supporting our key partners and customers. What has happened in the last 7 months, I would say, is an intermediate category has started gaining a lot of momentum. And that is either extra dense pluggables or pluggable CPO or near pluggable optics. These use cases bring the optics closer to the GPU of the switch ASIC but do not co-package them inside the ASIC or the switch. This allows for an easier deployment and a quicker deployment. And here, the need is 200-milliwatt laser singles or 100-milliwatt laser arrays, both of which we are sampling as we speak. So we are bringing forward the right products at the right time for meaningful engagement with customers across product categories and use cases. Let me dig a little deeper into the progress we have made with traditional pluggable transceivers. The 2027 revenue potential for pluggables growth, multiple customer engagements supporting the opportunity pipeline. With Jabil, alpha [ bills ] have been completed for preliminary system validation, beta [ bills ] in Q4 2026, followed by customer qual cycles. Initial product orders are expected in the first half of 2027, with the production ramp planned second half of 2027. With the Jabil announcement, several other pluggable manufacturers have meaningfully engaged with us. Three are in alpha sample evaluation stage, and another three in technical engagement and supply assessment stage. Some of these engaged module makers are capable of very rapid qualification and ramp, and many of the prospects continue to be identified. So this is very exciting as their contribution to 2027 revenue potential continues to grow. And also, this is validation of our opening up the aperture to target the pluggables market as well, which is severely supply constrained. More recently, we announced a $3.4 million new program with SemiNex, who is a long-standing customer of ours. They are receiving solid data center traction in co-packaged optics. And the initial focus is on semiconductor optical amplifiers. Why this combination? Sivers has 25-plus years experience in the design and manufacture of CW DFB lasers and semiconductor optical amplifiers. And we have a world-class indium phosphide technology platform for differentiating the site. SemiNex has 20-plus years of experience with high output power and high patient lasers and [ SOX ]. They have innovative low-loss waveguide designs and custom epitaxial designs along with chip-on-carrier package capability. That combination is very powerful to leverage into co-packaged optics within the AI data center. So stay posted on further developments over time on this partnership. Now I want to talk about the $4 billion SAM expansion for Sivers for our semiconductor optical amplifiers in optical circuit switches inside AI data centers. So switches you find in scale-out as well as scale up, why optical circuit switches, traditional switches are electrical or electronic switches. Optical circuit switches have lower latencies, lower power consumption and allow for decreased network complexity. So in the future, electrical switches will be used for fine grain routing and control, while optical switches will be used for heavy duty continuous training loans. Our [ lineage ] and SOA process technology and design experience is very relevant here, and we are now opening up the aperture to find more opportunities for our SOAs inside optical circuit switches. Now I'd like to move to a strategic shift in our manufacturing strategy in Photonics, where we are moving from a fab light to a hybrid manufacturing model. So what's the rationale? The strategic rationale is bringing in fresh capacity and developing global supply resilience. In July, Lumentum's CEO, Michael Hurlston, indicated that the supply shortage of indium phosphate could be even more severe than that of memory. Multiple areas for supply constraints were outlined. We continue to develop strong partnerships with indium phosphide substrate vendors as well as epitaxy vendors. But what we can directly influence is indium phosphide fab capacity for the laser chips and SOA stem cells. Laser demand outstrips supply next 3 to 5 years. Also, generational shifts in optical interconnects are happening faster. So CW laser demand continues to grow very rapidly as speeds evolve quicker and quicker. So this strategic shift is for bringing meaningful capacity online, leveraging our U.K. footprint in Glasgow. This hybrid manufacturing of Sivers and our foundry partners allows long-term supply resilience. So our positioning in the marketplace is that our lasers and SOAs are designed by Sivers Semiconductors out of Photonics Valley, U.K. with the tremendous experience we have across design, manufacturing and technology. But we are leveraging our foundry partners where we have added a new foundry partner who was brought on tremendous capacity that is available now to meaningfully address the persistent gap between demand and supply. Concurrently, we are expanding Sivers' manufacturing. Phase 1 expansion is underway in Glasgow already. And that additional capacity will be available from the end of Q4 2027. We are moving to a long-term capacity model where 1/3 of manufacturing capacity will be internal, while 2/3 will come from our foundry partners. Key takeaways, there are three. Number one, we are focused on pursuing the North Star. And that means making the tough near-term choices like Heine outlined to drive the three horizons that matter. The first one being Q4 2026 for quarterly revenue inflection. The second one being 2027 for product revenue inflection to lead us to the ultimate horizon, that's 2028 plus, where we are executing to our long-term financial model. Number two, transformational product ramps in 2027. Production orders continue to layer in, and we are consciously reallocating resource to deliver this key mission. This is the essence of the Sivers transformational journey, 2027 to drive inflection of product revenue versus NRE and bringing in the initial production orders that trigger multiyear production life cycles. And number three, a strategic shift to hybrid manufacturing, where we are bringing meaningful capacity and supply resilience to the AI data center ecosystem between our own expanded Sivers manufacturing footprint and our foundry partners bringing in tremendous amount of capacity that's ready now. Thanks for listening, and now we are open for Q&A.

So I'll take the first category of questions, and that's on the ops pipeline. And it's around are we seeing traction with NPO customers? How is our engagements with CPO and pluggables? As I mentioned in the call, we have engagements across all 3 areas of pluggables. We have opened up the aperture to work on pluggable opportunities as well as CPO takes some time to roll out. And we are seeing active engagement in pluggables and NPO, which are good sources of near-term revenue for us. So that's what we are staying focused on to make sure those ramp up. And then as more opportunities come through on the CPO side as well, we're well equipped by dealing with our partners and customers that are engaged with us to be ready for that as well. Heine, can you address a couple of the categories, please? Okay, let me address the question on capacity planning and indium phosphide shortage. So we have very good supplier partnerships, both on the substrate side as well as on the epitaxy side. As I mentioned, those are essential partnerships for us to make sure source capacity is available to us. When you get past epitaxy and substrates, the next thing is actual indium phosphide laser design capacity. And there, as we indicated now, our new partner has brought on board tremendous capacity for us. And therefore, we are able to ramp up very quickly on all the opportunities that are coming our way. So at this point, we are well equipped in terms of substrate, epitaxy as well as indium phosphide wafer capacity. And as I indicated, the long-term model is 1/3 of our capacity will come from within Sivers. And then 2/3 of it, we will have our partners address. We continue to look at the questions. So give us another minute or so. Heine, there are also questions about the NASDAQ listing. So do you want to give our viewers an update? [indiscernible]?

It looks like Heine has an audio problem is what it is, right? That's why the first minute after this call was quiet because Heine was answering a bunch of questions, but they did not come through. It seems he has an audio problem.

Got it.

So I think -- so Vickram, I think what we should do is just give it like 20 seconds because Heine is trying to resolve his issue on his side, and then we'll just go back to the question that he was reading.

Yes. I will address the other questions. All right. So we already talked about the pipeline between pluggables, NPO and CPO. And then I gave the audience an update on indium phosphide capacity. There was also a question on gross profits and gross margins. And I want to make sure that our audience understands. As we are consciously migrating from an NRE-based revenue to product-based revenue, we will have margin expansion. Our products business is definitely higher margin than our NRE business. And that's why we're taking consciously the tough steps now to stay focused on these production orders and these product ramps that are starting to flow in, and I showed you how they're starting to flow in very regularly now, and make sure that we enhance our gross margins by delivering more and more products. And the good news is all of these are multiyear production cycles. And as we get production orders that we satisfy for '27, we expect more production orders for the coming years as well because these are multiyear production cycles. Okay. There was also a question on, are you guys looking at M&A in addition to what you are organically executing? As a company, we are always interested in assets that may be out there, we cannot talk about those, but we make sure that we have pretty good activity looking around at what may be assets out there that are interesting for Sivers. Let's see here, that other categories.