EnSilica plc (ENSI) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen, and welcome to the EnSilica plc Final Results Investor Presentation. [Operator Instructions]. Before we begin, as usual, we would like to submit the following poll. And if you could give that your kind attention, I'm sure the company would be most grateful. And I'd now like to hand you over to the executive management team for their presentation. Ian, good morning, sir.

Good morning. Thank you, Jake, for the introduction. Welcome, everyone. Thanks for joining us for EnSilica's Financial Year 2026 results for year ending 31st of May 2026. As I said, I'm Ian Lankshear, the CEO; and I'm with Kristoff Rademan, our CFO. The 2026 results are really record set of results, really showing the business model in action and its ability to scale. Just a normal disclaimer about forward-looking statements in the presentation. Please read that at your leisure. Just for anyone not familiar with the company, a quick introduction. I mean, EnSilica, we're a fabless semiconductor company. We're experts in designing and supplying mixed-signal chips for high-growth markets. We were founded in 2001 as a design service company and have over the 25 years, have built a company a fabless semiconductor company. Now in 2021, when we started, there was an average of 300 chips per car, and now there's over 3,000 chips per car. And in that time as well, the complexity per chip has grown more than 6,000x according to Moore's Law. We're focusing on the space sector, but we also service adjacent markets that require high-quality, high reliability chips such as automotive and industrial. We do 2 types of chips. One is an ASIC, which is an application-specific integrated circuit. That's a chip that we -- is funded largely by the customer and is sold to one customer. That goes through the design cycle, and then we have the long-term reoccurring supply revenue. The second type of chip is called an ASSP, which is application-specific standard parts. These are our own chips designed to address a particular application. Our focus on particularly the user terminal space sector. These are sold to multiple customers and funded by a combination of lead customer funding, our own funding as well as managing to leverage U.K. space agency and European space agency funding. We balance between the 2 models with ASIC funding innovation and ASSPs because they're sold for multiple customers and have i.e. IP content, they're much more scalable with higher gross margins. In this business model, IP is very important. And as we go through designing the chips, we generate reusable IP, making every program faster and reducing the risk for the development risk. So if we look at our 2026 results, and our key achievements, they really do demonstrate the business model in action and working. So we had record results. So that was a step increase in revenue, leading to cash generation and really pleasingly excellent sales conversion and pipeline replenishment. Also, the business model is showing strong visibility of future revenue and profitability. We've got 5 ships now in long-term supply and 14 more queuing behind them. And we'll go through some of those in more detail later. Very importantly, we raised $25 million of new capital over 2 oversubscribed fundraises. And a big thank you to all of our new and existing shareholders that support us, supported us in those very successful fundraising. Those fundraisers were very much focused on growth, and that was growth both in the space sector, improving our EU footprint as people may have seen through the announcement of our Milan office and allowing us also to scale the supply revenue as the ships move through from design into supply and require that working capital. And also space is now a major tailwind for us. We have concentrated IP in this area. We're heavily engaged with most of the European and U.S. space operators. So that's a key market for us, and you'll see us very much leading with Space now. I mean these results have shown that we have a repeatable model here. Every program we have when it reaches production, it adds chips, volume supply improves our margins and our operational leverage without proportionally increasing in costs. I'll hand over to Kristoff just to go through the finances.

Thanks, Ian. So just talking through our financial momentum. We're obviously very pleased to report our results for financial year '26. With revenues of $27.8 million, they were up 53% versus the prior year, which was a great result for us. This flowed through to EBITDA of $5.9 million, up $5.9 million from the prior year, another great result and shows great progress versus financial year '25. Cash of $7.6 million, up 280%, and that is before the July placing of $14.9 million, which was oversubscribed. Then we come to the forward-looking indicators, which are of real strategic importance to us $375 million of lifetime supply visibility. That's 50% up on the prior year, and that's due to the contracts that we won during the last 6 months. Also looking at our pipeline, $600 million, that's after transferring $125 million into the lifetime supply. And this just shows that our opportunities are increasing as we become more established as a European player of note. As I mentioned before to other investors, we don't expect to win all of that $600 million, only about 25% to 33% over the next 12 to 18 months. But the main point is the opportunities are increasing, and we're looking to take advantage of more of those as we grow. Thank you. So just looking at the revenue split, chip supply revenues of $6.7 million, up $1 million on the prior year. This was a little bit disappointing for us as we had an issue with our U.K. automotive customer. They had a cybersecurity issue, which impacted volumes in financial year '26. They are now back up and running, and we are looking for volumes to go back to the pre-cybersecurity issue levels and then with modest growth in '27. Also, our industrial chip, which started in financial year '24 is continuing to scale up with volumes 15% ahead of the forecast. So really doing well. And for that reason, we expect growth of around 30% in financial year '27. And we also expect first orders from another industrial chip, our Edge AI chip and also the German automotive customers expected to go into supply in the second half of the year. So good growth expected in financial year '27 and then further growth in financial year '28 as these chips go into volume supply, and we should start seeing 12 months of supply from them as they scale up. So financial year '28 is expected to be even better than '27 in terms of supply growth. Then moving on to the NRE, $15 million versus $5.9 million, a very large increase. '25 wasn't a particularly good year for us in terms of NRE, but we've bounced back really well in '26. And in '27, we expect similar levels, maybe with a 10% growth on that. So some growth, but not a huge growth. And this is built based on the 4 to 5 new contracts that we expect to win every year. So we keep that treadmill going, and we see ultimately the real growth in the supply as these chips that we're developing go into supply. Consultancy revenues, whilst the key area for the company, the employees of the business are being pulled into the NRE and development contracts, which means that the focus is not totally on the consultancy. So we expect that to be stable, but really the growth will come from the NRE and ultimately the supply revenues. Thank you. Then looking at the income statement, we spoke about the revenues. Gross profits has gone down from 40% to 35%. That is because we are investing heavily in our ASSPs, as we said we would as part of the fundraise. And effectively, the development costs relating to that go into cost of sales, whilst the income from that is sitting in both revenues from our European satellite operator customer and then also in other income as grant income from the UKSA. And if that grant income was included in revenues, then the gross profits would be at a similar level to what they were in financial year '25. And I guess the point we're trying to make is that we're not doing worse than we did in '25. It's just that the income is sitting in 2 different lines rather than only in the revenue line. We also had a benefit from the R&D tax credit being higher than initially assessed, and that has flowed down into EBITDA of $5.9 million, which is significantly up from the prior year and a very good result for us. We're also pleased that, that has flowed down into an operating profit of $2.9 million for the business. Thank you. Looking at the balance sheet. Intangible assets have increased to $27.3 million. That is as a result of our continued co-investment in the development of ASIC and also our own IP and know-how. And that is underpinned by the $375 million of lifetime supply that we spoke about on the previous slide. So we do assessments of each one of the intangible assets every 6 months and where necessary, we will impair or write down the assets if we can't -- we don't expect to recover the value. But as you will see, it's more than covered by the $375 million. So that is looking very, very positive. Cash we spoke about. And obviously, that is before the benefit of the second $15 million placing, which came in post year-end. Loans, the balance has come down as a result of repaying the loans. We don't expect to repay the loans earlier than the repayment terms, mainly because we did the fundraise in order to fund our growth rather than to repay our financing. So our focus will be on growing the business, investing in the space and SATCOM's ASSP in order to really bring forward our revenue targets for the short to medium term. Then looking at the cash flow, a very successful year for us with cash consumption of $1.4 million, down from $5.3 million. That was as a result of the moving forward the development contract, getting the upfront and milestone payments and then moving those forward. Also the RDEC tax credit, which increased slightly. We expect a larger increase in financial year '27 from the FY '26 claim. So that should be positive. And then intangibles and CapEx investments, slightly higher than the prior year, but not significantly so. What I would say is that for financial year '27, there will be an increase in cash consumption. This is in line with our stated intent as part of the fundraise that we intend to invest in our ASSPs in order to generate larger future revenues. So in '27, we see cash consumption go up alongside a bigger investment in intangibles and CapEx investment. And then in '28, that decreases again mainly due to the increase in supply revenues that we expect in that year. So towards the end of the year, we expect to go cash flow positive on an operational basis on a monthly basis. And then financial year '29, we expect to go full year cash flow positive and on a free cash flow basis. And that is mainly because of these chips that we are developing now that are going into supply and will be fully into supply in those years, and we should see a significant increase in the level of supply revenues from where we are today. Thank you.

Thanks, Christoph. I mean I'll walk you through the strategy and operational review. I mean just to remind people what a fabless semiconductor company is. I mean I talked about it earlier. I mean a fabless company as it sounds a company that doesn't have a wafer foundry. It leverages wafer foundries, such as the largest one, the TSMC and GlobalFoundries. And silica is in a position where we have direct relationships with those major foundries. So it uses a capital-light model. We don't have fabs and the other part of the supply chain is outsourced assembly and test houses that would take the wafers, dice them up, package them and test them under our operational team management. So this is capital light. It's fast time to market. We're able to leverage the latest technology, proven technologies and allows us to really focus on the IP and innovation and the customer requirements. Okay. This is a relatively new slide. People may not seen it, and it's really to demonstrate the development cycle we're going through and how much -- it's very much compounding. I mean we -- in the ASIC model, as I mentioned, that's largely funded the NRE by the customer. We go through the design and then we reach the key milestone of the tape-out of the device where we send the masking information off to the foundry to fabricate the masks and then the wafers. And then we get the chip back and we go through the validation of that. Then it then goes into volume supply, so we get the long-term supply revenue. And that's often 7 to 15 years in the markets that we're addressing. But the other products of the development is we get reusable IP and know-how that goes back into our toolbox, and we can reuse that from chip to chip. So every -- it's like a snowball effect. Every time we're going around this, we're generating IP, the next one becomes easier, we're able to increase our gross margins as well. So on the ASSP, that's a sort of similar cycle, except for the development costs are funded by in silica. And we bring in lead customers early to support that as well as other funding such as European Space Agency and U.K. space agency funding. So we go through the same cycle of developing the chip. Taking it out validation. And then that chip sold to multiple customers, meaning it's much more scalable in terms of larger number of chips can be sold, higher margin. And the other advantage is once that chip is in production, any new adopter of that chip, we can get an uptick in revenue relatively quickly. I mean they take the chip, they evaluate it, put it in their product. We don't have to go through the design. We don't have to go through that design cycle as we have with the ASICs. And also out of this, we get IP. It's often higher-value IP in terms of platform level IP as well as patents that go back into that toolbox that we can use either in the ASIC model or the ASSP model. So I mean very much is a sort of compounding cycle we have. We -- when we started off in this business model in 2016, we virtually had an empty toolbox there of IP. And now we've got strong IP. We've got customer references, and we can demonstrate that we're a reliable designer and supplier of chips at scale. So this portfolio here really does demonstrate the business model in action. I mean we have 5 chips now in supply, generating that reoccurring revenue we talked about with leading with the automotive one, and then we have a very strong industrial one in #4 there. And they're in supply and ramping up -- still ramping up strongly. And then we have another 14 chips in the design phase. I mean this year, our team -- our design team and operations team have executed fantastically. In '26, we had 2 tape-outs. So far in '27, we've had 3, and we have another 2 schedules. So those chips are moving forward into the supply area now. So just sort of talking about some of the color codes here, where you see the gray colors on the squares where you've got year 1 to 10. There we've got forecast from the customer, and we have revenue forecast for each of the chip and the likely lifetime of the chip. And if you take that and reflect that forward on the revenue illustration on the right, you see the dark blue, which is the growth in the revenue that we're seeing from the chips with confirmed forecast. So although it's colored dark blue, it's not blue sky. I mean these are real customer forecasts, and that's real growth that we're seeing there in the supply revenue. Those orange ones that you see, which are largely the space ones are our ASSPs. And you'll see there that 8 and 15 are the ASSPs where we've got confirmed design-in slots with our customers. So just turning to the graph on the right, just explaining the other elements. So what you see here at the bottom in the orange is the -- the NRE revenue, and that's driven by the 4 to 5 ASICs that are feeding the pipeline. And they're not actually -- the revenue from those ASICs is not shown in the supply growth there. They'll be the new design wins that you should be looking out for next year where we're targeting 4 to 5 ASIC design wins. We talked about the consultancy revenue, and there's no plan to grow that, but we have a steady layer of consultancy revenue. And what you see here on the light blue side at the top is a conservative estimate for the growth in the ASSPs. And their space ASSPs for user terminals, and we see them really coming online in beyond sort of 2030. And I'll cover the space sector in more detail. You'll be able to see from that, there's loads of upside potential on those. So now if we turn to the space opportunity, it's very much a multi-decade opportunity. You can see there. I mean, there's huge growth. And one thing you can say about Elon Musk is that one positive thing, this is really proven that space-based communication works and it can be delivered at scale to a huge subscriber base. I mean he has over 11 million subscribers now and really shown that resilient connectivity is required. And that's why there's lots of sovereign and other constellations coming online really to deliver similar services. They're enabled by falling launch costs. These mega constellations require lots of satellites. So it's important that they've got the launch facility. So it's not only SpaceX there, you've got Blue Origin, you've got Rocket Labs. So there's a number of alternative services there for launching the satellites. I mean there's also an exploding demand for bandwidth as data need to go back to the cloud for processing and generally, the requirement for high bandwidth ubiquitous connectivity. And this is all driven by strong government and private investments. I mean you would -- you see that earlier last week that SpaceX announced a $40 billion fund raise and a lot of the other satellite constellations are raising billions. It typically takes $5 billion to $15 billion to put up a constellation. I mean as a chip company, I mean, we always look for volumes because once we ship the volumes, it keeps the foundries happy. And we really see there the user terminal area, which is the piece of equipment, consider it like a set-top box, which is used to connect to the satellite and give you the connectivity to the constellation. I mean that is -- that's where our IP is concentrated, and we're focusing our efforts in developing a family of ASSPs, a full chipset, and I'll talk through that in more detail. I mean you should be looking out over the next 18 months for us and announcing more ASSP design-in wins with major constellations. So as well as that user terminal chip area, I mean, we are also developing chips that go into the satellite payloads. So the satellite itself carries a payload. It's called a communication payload. And within that payload, there's a very complex chip called a beam former. We're experts for that beam former. I mean we announced that back in 2020, we were developing AST Space Mobile payload chip. I mean these are very high-value chips that form beam, so you can change the coverage profile of the satellite. So for instance, as you're going across the ocean, you don't really need to give coverage for an empty ocean. You may provide a beam on or areas that have poor coverage. So as I said, this is a complex chip, very high value, and we've got world-class expertise on that. So these ASSPs, I mean, we have 4 of them now in silicon form, which are being evaluated by different customers. We're engaged with 3 major operators. I mean we announced a $50 million design in of our -- one of our user terminal ASSPs with a major European satellite operator. And we've been funding this both from lead customer funding and space agency funding. So we've had over $30 million of funding to support the ASSP development in this area. So if we just zoom in a little bit more about user terminals and user terminal ASPs -- so on the left and right here, you can see some typical user terminals. There's a high-end one on the left from one of our German customers, VITES, which is used for enterprise type applications. And the one on the right is a mobile terminal. The great thing about these, particularly for semiconductor company is they contain many hundreds to thousands of chips. They include RF millimeter wave chips. They include a beamformer chip and also a modem chip. And we are one of the only company that is developing a full end-to-end chipset with reference designs to enable low-cost highly integrated user terminals, which are rich in differentiated features. I mean this allows our customers to have a fast time to revenue, enable them to have low-cost user terminals so they can sign up the user base, which is very important in terms of getting the satellite terminal revenues up. I mean it's worth saying at the moment, this is dominated by Starlink and the user terminal market for semiconductors is around $650 million. You've also got Amazon, which is rolling out. I mean they have an initial constellation there with rolling out some user bases and Eutelsat who now own OneWeb have a LEO constellation. But as we move through, there's deployment and testing going on for additional constellations such as ViaSat and Blue Origin who's known for their rockets have one called TerraWave. Telesat is a Canadian constellation. They're rolling out services and be signing up for users. And then really 2030 is when it happens where that market grows to over $3 billion, where you've got Iris Squared, which is an EU sovereign constellation and the likes of SES as well as enterprise and defense and sovereign constellations. I mean, for example, the Chinese car manufacturer, GE is putting up their own constellation. So really, this is opening up the market of having user terminals embedded in cars similar to equals and gives us even a bigger automotive TAM with something like 75 million to 80 million cars still being in production. So a great market. We've got differentiated IPs and are developing a highly integrated chipset to enable those satellite operators, which are not vertically integrated to provide state-of-the-art user terminals to rival those of Starlink. So if we talk about adjacent markets, automotive and industrial, it's a key market. It's where we cut our teeth. It's all about high reliability, long-life mixed-signal chips. I mean we're applying our know-how and ability to deliver at scale into that market into the new space area. But we're also converting very good automotive and industrial design wins. As people are aware, we have a $40 million design win for -- with the U.K. OEM, where we've shipped over 10 million chips into that area now. We recently announced a $75 million design win with a German Tier 1. And there's other very strong design wins where we have, for instance, Siemens as a customer who is one of the leading industrial OEMs. So a great market that produces long-term reoccurring revenue, almost like an annuity scheme. So still a good market, still a growth, and I talked about the number of chips per vehicle and per production line increasing massively. And if we just turn to Photonics, Kristoff really likes this market, and I'll let him describe the opportunity we have.

Thanks, Ian. So Photonics, as a lot of you might know, make use of lights instead of electrons, which deliver high speed, lower power and less heat than conventional chips. So a huge opportunity there. Where in silica specializes is photonics controllers. And these are the chips which control the photonics chips and the photonics rather than being photonics chips themselves. We announced the contract in 2025 with Oriole Networks, and that chip is progressing very well. It's for a photonics controller. And Oriole Networks recently announced a tie-up with AMD, and that gives us a lot of hope that their product will be even more successful than we expected it to be and that they will go from strength to strength. We are also seeing interest from other photonics companies and also quantum computing companies for controllers in that area. These are quite similar and make use of the same engineers to develop those controllers. So we would hope that in the next 12 months, we'll be able to sign up further customers in this area. We see this as a high-growth area and could turn into a megatrend in the future. And we are looking to develop this area and grow in this area, although our main focus, as Ian highlighted, is really in the satellite and communications area, where most of our development and investment will take place, but we will do some selective investments in developing our photonics.

Thanks, Kristoff. I mean just to point out, I mean, this is an area because of the diversity of the approach to photonics, this is an ASIC area rather than an ASSP area. But very exciting. We really -- these will make a big difference to the AI hyperscaler data centers, reducing the latency between the multiple processes they've got, making them much more efficient and reducing the power. So I mean another area which really is used across all of our markets, particularly communications, industrial and automotive is that of safety and cybersecurity. And really, there's a change in legislation, the U.K. EU Cyber Resilience Act and equivalent ones in the U.S. that really are increasing the baseline of security needed in semiconductor devices. I mean in silicon, we have our own IP, where we've been developing both classical and post-quantum cryptography and security subsystems. And these are now being deployed in our ASICs and ASSP devices. And really that we're turning what is a regulatory requirement into a competitive advantage by leveraging that IP.

So just to summarize, space is a key growth area for us, and we see this as a megatrend. The user terminal market where we have a lot of our IP grew by 5x by 2030, and we see this as a major tailwind for EnSilica. We are investing heavily in order to have our chips ready for when the European satellites go up so that they can go into those terminals and support our customers. And we feel we're in a fantastic place to take advantage of this growing market. In pipeline, we continue to target 4 to 5 design and supply contract wins annually. So please look out for new design wins over the next few months as we execute on our stated strategy and business plan. Momentum, as Ian mentioned, we've had 2 tape-outs in financial year '26, 3 in financial year '27 and a further 2 tape-outs expected to complete in the rest of the year. And as explained, this is a good forward indicator that these chips will go into volume supply in the next 12 to 18 months post tape-out. So it really shows that these chips have moved forward and are much closer to volume supply and good progress seen. Margins our chips are moving into generating recurring supply. That's driving up our EBITDA margins, as you will have seen from our results for financial year '26. And we're really looking to grow to our target of a 30% EBITDA margin on our revenues. And we would hope to be able to achieve that target in the short to medium term. And then guidance, we are reconfirming our target revenue of $32 million to $34 million and EBITDA of $5.5 million to $6.5 million. Thank you very much.

I may just jump back in there. Thank you very much indeed for your presentation this morning. [Operator Instructions] Just to remind you that a recording of this presentation along with a copy of the slides and the published Q&A can be accessed via investor dashboards. We have received a number of questions, and thank you to all of those on the call for taking the time to submit their questions. But Ian, Kristoff at this point, if I may just hand over to you to address those where appropriate. And if I pick up from you at the end, that would be great. Thank you.

Thank you very much. So running through the questions, do you envisage a future contract for your photonics controller ASIC with either Ori Networks or AMD? And if so, knowing the lead up time to a volume could this be this year? We already have a contract with Oriol Networks. We have no tie-up to AMD. It is Oriol that announced the tie-up with AMD. And unfortunately, we're not able to discuss the detail of that specific contract in terms of when the volume rollout will be. But obviously, when it does go into production, we will notify the market of that even if we're not able to specify the customer. Are you in discussions with any other companies regarding your photonic controller ASIC? I think as we explained for ASICs are products that are for specific customers, and therefore, we are not able to sell it to another customer, but we are speaking to other customers about doing ASIC photonics controllers for them and also potential quantum computing customers, and we will obviously announce those should we win those contracts. Could your photonic controller ASIC be used to switch qubits in quantum computers with minimum alterations to the blueprint?

Okay. I mean there are different types of qubits and the way they're excited changes. So you've got some that would use sort of photonic methods, some would use RF methods. So we've got applicable technology to make controller chips for those types of qubits.

With AST Space Mobile launches in full swing and FCC approval for the expansion from 48 to 204 satellites, when do you perceive you will get a revamped order for the next 198 payload ASICs known as the AST 5000 ASIC?

Okay. I'll pick that one -- so we did an announcement back in 2020 about working with AST. But unfortunately, we can't comment on AST's launch profiles or potential orders.

Given the comments around the substantial growth opportunities for ASSPs, the intention to use part of the part of the capital raise to bring ASSPs to market and the expectation that ASSPs will be an important driver of your revenue beyond the $100 million target. Could you please give us a little more color on the commercial progression you expect to see? In particular, without providing formal revenue guidance, could you indicate roughly outlined pathway from a new ASSP design win to commercial production, e.g., the typical time from design win to volume shipments, the approximate number of units that successful design could represent and the sort of ASSP or revenue opportunity per design win.

All right. A very detailed question. But I think we sort of demonstrated before with that previous slide on Slide 14 about the growth of the ASSPs and the fact is that the growth will start in 2030, and that's really to do with when the satellites are up and they're building the subscriber base that means they'll be ordering the user terminals. I mean you can -- I've talked about the many chips that you have per user terminal and really a new constellation. As I mentioned, they're going to be between $5 billion and $15 billion to do a full-blown LEO or LEO constellation. So they will have many tens of thousands to -- could be up to millions of subscribers and each one of those would need a user terminal. So I mean, we saw an announcement previously of designing of one of our chips, which was 50 million. I mean we see that such design-ins will be worth similar amounts, if not, they could be in the hundreds of millions. But really, I said that will all come when the launch starts and the services production the services online. So really, this is a longer-term revenue out there. We'll be getting some in terms of test terminals in '29, but really will be 2030 and beyond.

Satellite networks are increasingly exploring optical links between spacecraft and between satellites and ground infrastructure. As in silica identifies photonics as a strategic growth area and bearing in mind the investment made in the photonics controller program, do you see space-based optical communications as a potential future revenue stream?

So I mean very good observation there. And I mean, satellites are increasingly using optics to link the satellites together, that reduces the number of the number of ground stations because they can have a fast link between them and they then be the data down. And it is very -- again, good observation is that the technology we're dealing in can be applied to those optical links. So as well as being able to use our millimeter wave experience for those high-frequency links, we can also apply that to the optical links and the chips that are used to control those links.

Of the chips that have reached or are expected to reach tape-out in financial year '27, what is the typical lapse time from tape-out to first customer revenue and then to volume production? So it's normally around 12 months, but could even be as long as 18 months. This is because the chip takes out and goes -- comes back to us, we assess it. It also goes to the customer. They have to assess and make sure it's fully built into their product, which takes some time. But normally, there's a push from both sides to get it in as quickly as possible. So could you please provide an update on the SIA Micro contract given that SIA entered extraordinary administration in Italy in May 2026. In particular, has -- in silica had any further engagement with SIA or the appointed administrator regarding the outstanding $1.6 million provisioned in financial year '25, the status of the EU IPCO funding and the potential resumption or transfer of the ASIC program to another party. How do you assess the prospects of recovering the outstanding debt and any future revenue from the contract given the change in SIA's financial and legal circumstances? We fully provided for the debt. We've taken SIA out of our forecast -- we don't, at this point, expect the program to restart in the short or medium term. So -- and we can't really comment on how likely it is that we will recover the debt that we provided against. So the best to say is that we've moved on to other customer programs, and we're not currently engaging on a regular basis with SIA. I refer to your Easy Crystal 3-in-1 QQC accelerator in a single hardware block -- given that NIST is now actively encouraging migration to PQC, how significant do you believe that the architectural advantage of your implementation is versus competing QQC IP? Do you see potential for the easyCrystals architecture to become broadly adopted IP implementation across multiple semiconductor designs? And are you seeking evidence of that in customer evaluations or licensing discussions?

Okay. Very good observation there. So the easy crystals basically takes the PQ algorithms that 3 and 1 and have them integrated together so they can save silicon resources, save silicon area. And we see that very differentiated in terms of allowing you to put PQC in your chip without massively increasing the cost of the chip or allowing you to target more mature nodes. So we're building that IP into a number of our own ASICs and ASSPs, and we consider that to be a real differentiator for us. We do offer it for licensing. And I mean, we announced one license to the IP, and we have some evaluating it. But the prime purpose of the IP is to differentiate our products and use it internally.

With regard to the PQC IP, Cherry and RiskV capability together with the 5 million U.K. government-funded secure processor program, could you please explain how you expect this to translate into commercial revenues over the next 3 to 5 years? Do you see Cherry PQC as a differentiator that helps win larger ASIC contracts or as a potentially significant stand-alone IP product business? When do you expect the first meaningful revenues from it?