SolarEdge Technologies, Inc. ($SEDG)

Well, great. Thanks, everybody, for coming in bright and early. Very pleased to have Shuki Nir here, the CEO of SolarEdge. I'm Jeff Osborne, the TD Cowen Sustainability analyst. Thanks for sticking with us on day 2. Shuki, thanks for taking time out of your busy day to join us and being here from Israel. But I imagine most in the room are familiar with SolarEdge, but just in 90 seconds, a minute or 2 to go over what SolarEdge is about in terms of the platform you've built would be helpful, and then we can dig into some questions.

Thank you. Thank you for having me, and it's great seeing you here and online. So yes, SolarEdge, this is our 20th year. The company was founded on understanding power electronics in general and DC in particular. And based on that, over the years, it built different products and different solutions that are actually bringing value to the end customers, whether it's the homeowner, C&I, now we are going to talk about data centers as well, AI data centers in the future. And what we are trying to do always is to lead by the innovation, the quality of the technology that we develop and make it into workable solutions that are actually increasing the ROI on the project or on the solution that we are providing. That will be the 90 seconds.

Perfect. I appreciate it. I definitely want to touch on data centers. Maybe we'll start there, but I promised to ask quite a few questions on both storage and solar as well. But just given the explosion of interest among the buy side in data centers, maybe we'll start there. So maybe at a high level, let's just walk through how the relationship with Infineon is proceeding? And then I believe the goal is to have sort of alpha and beta units next year. Right now, both yourselves and many peers are sort of in the white paper marketing, waiting for the NVIDIA product cycle to play out then with pull market adoption. But maybe just let's start with the Infineon relationship, what the history is there and how it's progressing?

The history with Infineon is a long -- maybe since almost the beginning of the company, we partnered with Infineon. They are very, very close partners of ours. The silicon carbide and other solutions that they offer is part of our inverters today. And when we announced the entry into the AI data center area, we said that we announced it together with Infineon because the silicon carbide that they are making is going to be part of this product. And the partnership is progressing well. You mentioned that we are in the white paper marketing area. This is not exactly true about SolarEdge. SolarEdge, about 5 or 6 years ago, we started developing -- back then, it was aimed at the utility space, solar utility, a solid-state transformer. So for 3 years, the company invested a lot of money in creating the infrastructure and the labs and everything that is needed and dozens of engineers had actually worked on it. Until 2024, the company had some difficulties and because of expense control, we put the project on hold. Last year, when NVIDIA released the white paper talking about the expected and very much needed transition to DC infrastructure in the data centers, in the AI data centers, we went and we basically brought this project back to life. This is why we keep talking about the fact that we have all the building blocks already and now we are putting them together. So we put the team together, brought back actually some people who left SolarEdge because we closed that project. And we have some very, very talented people who understand DC for the last 20, 25 years, which is an area of expertise that is very much needed. We are now at the stage that we aim at having a working product in our labs this year. And we are going to have pilot installations next year. And then 2028 is when probably that will meet the transition that NVIDIA is talking about in terms of the GPU road map that will require the transition to the 800 volt. So that's kind of how we look at it.

That's helpful. Just as a point of intellectual curiosity, when the old team pre-'24, was that at a 1,500 volt, which is the solar voltage for utility scale that then you're sort of dialing back to 800 volt. A lot of people would say it's easier to go backwards on voltage than it is to go up.

Yes. So I don't want to say nothing is easy in these technologies, but moving the solid-state transformer from being the solar product that it was originally planned for into the data center SSD, the latter should be less complicated and less complex. But there are many, many learnings that we've had. These things are not working on trial #1, right? So we had to go and learn, make some changes, make some adjustments over the 3 years that the development has been in place. And now we feel confident that we have the architecture in place. This is why we are talking about the 99% efficiency, while others are talking about lower efficiency. And we feel that there is a lot of work that has to go into it. Don't get me wrong. But the feedback that we are getting from the ecosystem, from the potential partners as well as our own feeling is that we are making good progress.

Got it. Maybe just staying on solid-state transformers, as we commercialize the product, what do you feel the top 3 KPIs will be beyond efficiency? Most people are in the 98% realm, you're in 99% today. But I think there's certainly -- the block size is different for everybody. I think your marketing suggests 2 to 5 megawatts. Heron Power, I think, has a white paper out at 4.2. Folks at Enphase are 1.25, but combining 342 small devices to get there. But what -- like as you talk to people, what's the right size if a solid-state transformer to come up behind you and bite you? What -- how big do you expect it to be?

I think that the size or the number that matters here is not whether it's 2 megawatt, 4 megawatts or 5 megawatts. The idea is to start from 34.5 kilovolt, the AC. This is the AC, the medium voltage that is coming from the grid. And you hear about different people that are having 10 kilovolt or 15 kilovolt, based on what we understand, at least. This is not meeting what the expectation is. The expectation is to go in one step from 34.5 kilovolt down to 800-volt DC. And it's complicated. It does require innovation, technology understanding and execution. So the first thing that I believe people want -- you mentioned efficiency. So it's -- the efficiency is also related with the fact that it will be one step from 34.5 down to...

So yours is one step, peers are multiple steps and that has a degradation.

I don't want to comment about peers, but what we are doing is that we...

That gives you confidence to be at 99% versus others maybe you're theoretically at.

Exactly. That as well as other innovations that we have got. Yes, absolutely. So that's number one. The second thing that is in discussions with the ecosystem that is important for them is actually the scale. We are talking about next year, we are going to have some pilot installations. That by itself is going to require some manufacturing capabilities. But if you go to 2028 and beyond, you're talking about the need to scale up manufacturing. The need to actually take it to the hundreds of megawatts, if not gigawatts of manufacturing. And this is something that the hyperscalers or the neoclouds, they are looking at it as this is table stakes, right? You cannot -- if you can't deliver on that, don't even show up. So that's the second thing. The third thing, which is very important for everybody in the ecosystem is to see that you have partnerships across the entire powertrain. So it's not a stand-alone product. It's actually part of a solution. As a data center builder, you want to make sure that everything is working together, playing together in order to benefit.

So it's construction at the front end, but what about like power electronics distribution? Do you need to interoperate or have...

So the product that we are developing is starting from -- on one end, is the connection to the medium voltage. Then you have the SST, so it's doing the transformation from the 34.5 to 800 volt. And then you have the PDU, a power distribution unit. But part of the change that is going to happen, as an example, is data centers today, they are using UPS that is connected to the AC. If you change to DC, you need to change the UPS infrastructure. So we are developing the DC UPS of the future, if you will. Whether you're a customer or a horizontal player in the market, you want to make sure that everything is talking to each other and you have a complete solution and not something that is just we develop -- somebody develops something in isolation, they bring it into the table and then it doesn't interact or doesn't connect efficiently with the rest of the power electronics infrastructure in the data center.

Got it. And then do you have any preliminary steps around go-to-market or the testing that you're doing in discussions? Or are those all with your own sales force directly today?

So the good news here is the number of players is quite -- it's not huge. It's -- we are used to go after 10,000 installers and after -- in the solar industry. Here, you're talking about several dozens of potential customers and the leaders are probably less than 20 people or less than 20 companies. We've engaged with most of them in discussions. And the discussions today are very, very technical. There are less -- it's not about go-to-market, pricing, quantities, et cetera. It's about the technology. From their perspective, they want to make sure that the solution they choose and pick is something that is reliable. I didn't mention reliability earlier, but obviously, this is a table space, right? Reliable does what it promises to do and delivers on time. So at this stage, we are discussing with them. They are trying to understand the competitive advantage that we have. And the go-to-market from then on, we don't look at it as something that is too complicated. You and I, we talked about it before. There are -- one way is to go directly to the neocloud and the hyperscalers. The other one is maybe through some collaboration with the power electronics providers of today. We know that they're trying to develop their own solutions. For them, it would make sense also if there is a better solution than what they're developing to try to collaborate. So we are talking to everybody. We don't have any exclusive go-to-market strategy that we're going to exclude any other thing.

Maybe last question for myself, and then I'll open the audience for data center-related questions, and I definitely want to pivot to solar in the last 15 minutes or so is just as I think you folks are aiming to have an Analyst Day later in the year and as investors try to size the market. My understanding, and correct me if I'm wrong, is that not only for argument's sake, we say there's a gigawatt data center that is moving to solid state, but then there's an element of reliability that you need five nines. I had to ask the Google machine what five nines is, but I think it's 5.26 minutes a year is what the 0.0001% downtime is. So expectations are obviously very high at five nines. But -- so is there an element of 20% overbuild just from a reliability perspective or 10% that I assume. So the gigawatt is really 1.2, 1.1 something higher. And then there's -- you need solid state. I think there's been an explosion of interest as we move to more on-site generation, whether it's gas turbines, reciprocating engines or fuel cells, just the power quality is pretty crappy coming out of those. And so there's a lot of discussion of putting batteries sort of between either the grid or on-site generation in the building, those would all need solid-state transformers as well, right? Or...

So you're taking it 2, 3 steps forward, which is great. And we're exploring these opportunities as well. We're in discussions with different . But at this stage, the focus of the team is to provide the SST that the data center, it's going to be a must, right? The next, next gen of the GPUs has to be supported by the 800-volt DC. That's what NVIDIA is saying. So the focus of the data centers as well as ours is to make sure that they have a working, very efficient solution available for them. Reliability, we're testing different levels of redundancy within our SSD. By the way, the products, some people think about product as something like that. It's a 20-foot container, right? So it's a large product. Our infrastructure is that these are 30 cells that when combined together, they are 5 megawatts. Inside the container, we are going to put some redundancy as well. And I believe that what you said is right, I believe that the data center owners or builders are going to put some additional redundancy on top of that. So we're expecting all of these things to happen. At this stage, next year, we are going to start working with 20, 30-megawatt installations. And then we're going to have much more data in order to be able to comply with the five nines or the seven nines. Some people are talking about seven nines.

Okay. Perfect. Any questions from the audience on data center-oriented activity? Otherwise, we'll pivot to solar. All right. Maybe just 30,000 foot and we'll drill down. But like what are you hearing from your channel partners as it relates to like the voice of the customer in solar, both maybe compare, contrast Europe versus the U.S. around interest in the category, both residential and commercial.

So I think that overall, globally, by the way, everybody understands that electricity prices will continue to go up. And solar is probably the most affordable and the best solution for that. So the underlying demand and the need for additional solar solutions, especially now when it comes with storage. So it's a solution. It's not just solar, we're going to produce more energy from the sun. Actually, we can now store it, use it later. Time shifting and optimizing the rates here is something that is very, very interesting for companies as well as individuals. So we continue to see a lot of interest and business. Now talking about the interest is one level. The other level is obviously the actual demand. What we've seen in Europe, and we talked about it starting in March, they really -- it came close to home, I would say, for the Europeans that electricity prices are about to go up, some dependency or independency of the grid or energy. So we are seeing an uptick in Europe that is beyond or above the seasonal uptick that we have seen. That continued into April. And from our initial checks, it continues also in May. So the Europe -- on the European side, both on resi and C&I, we're seeing this uptick. In U.S., and we talked about it before, the C&I market is doing well. We are very well positioned in that market, and we're seeing good progress over there. On the resi side, due to the lack of clarity around PFE and FEOC and slowness in the investment into the TPOs and the elimination of 25D, so that market has contracted, as you know.

You think we're at the bottom? When that recovers is a debate that...

Yes. So I feel that what we're seeing now is a slower level of investment or a lower level of investment in the TPOs. The bottom will be determined by when the level of investment in the TPO is going to go up. They will be -- that will be the catalyst for additional installations on this year on demand. And we'll see what we expected to see this year that, yes, cash and loan is going to go down. But the rest of the business, there is a lot of variability around this business today.

Just a follow-up on that is while the investment is lower, your visibility into the multiple years ahead because of safe harboring, I imagine, is better than in the past, like 2, 3 years ago, pre-Trump, you probably would be saying -- scratching your head saying, we have no idea when this is going to recover, but at least you have an alignment with the channel and partners. Is that the right way to think about that?

Yes. So we've -- over the last several quarters and actually continuing now until July 3, we've engaged in multiple safe harbor deals, both with the TPOs as well as on the C&I side. And what these safe harbor deals are going to provide us and the customers is better visibility to your point, into future revenue or future demand into a more aligned manufacturing and production planning. So from our perspective, it's all goodness. When exactly they're going to start pulling the safe harbor deals is depending upon customer preference, which is good.

Can we just pivot to the Nexus product launch? I think you started in Germany and then broader Europe, and then I saw on LinkedIn this morning that there's now training sessions in America to ramp that up with the channel. But what -- how do we think about where we are in that journey, the single SKU transition, all of those should have a pretty profound impact on margins and cleaning up the channel around.

Absolutely. Look, Nexus, I'm so excited about it that I can talk about it for the next hour, but I'll try to be concise. It's the platform that SolarEdge has launched that is addressing today's residential market. And I'll explain. It's moved -- the market has moved from PV only to PV plus storage. And it's coupled with the fact that the grid is becoming very dynamic. In Europe, you have dynamic rating every 15 minutes. Here in different states in the U.S., it starts to change as well. So what we are seeing is the need from a homeowner perspective, to get a system that doesn't only produce more power from a given set of panels, which SolarEdge has done for the last 2 decades, but actually that uses intelligence, artificial intelligence to optimize when to import, when to export, when to activate some loads and so forth. So that's one piece that Nexus is bringing to the table. The other piece, and we talked about it is it's very, very installer friendly. When you combine these 2, and we had the launch event back in March in Germany, the excitement from the installers was something that we didn't really expect to be that -- to see that level of excitement. And what we have is the entire Q2 supply that we have for Europe is already booked. We continue to have bookings for Q3. We are ramping up now the manufacturing. As for the U.S. market, we have pilot installations in the U.S. already. We did -- our guidance to ourselves was we prefer reliability and quality over time to market. So we've had dozens of installations here in the U.S. in order to make sure that the installation process is working well, the commissioning as well as the actual production and activation of the systems. We are going to start rolling out Nexis in the U.S. in the next quarter. And we're going to do it together. The market here is slightly different, as you know, with the TPOs and the different customers. So we're going to roll it out based on customer preferences and together with them.

Can you just touch on at a high level, the financial ramifications of single SKU and Nexus relative to the contract manufacturing footprint and it should be pretty accretive like more...

So there are 2 things that Nexis brings to the table from a cost structure perspective. The first one is the fact that it's made in the U.S. So all of our Nexus inverters and optimizers are going to be made in the U.S., which obviously is entitled to the 45X credits that give us a net cost -- a lower net cost. Nexis itself, apples-to-apples compared to the other or to our previous platform is lower cost. So we have better cost by definition. And then the multiplier is the fact that it's going to be made in the U.S. And the third point around Nexis is we believe that the value that we bring to the homeowners is significantly better and higher than what we brought in the past and what other people are bringing now, which we believe will give us a little bit more pricing power as well, which also contributes to margin, as you know. So between all of these 3 things, we're excited about Nexis, about the value that we bring to customers and the margin impact.

You think by next spring '27, the full transition will be done for Nexis in terms of that ramp?

Yes.

I assume it takes 3 or 4 quarters.

The goal is to do it as quickly as we can while maintaining reliability and quality. And the plan is by the end of Q1 to finish the transition.

Can we just switch gears to Europe -- back to Europe, but there's been some headlines with Italy having an auction that ban Chinese inverters. There's been some stories in Der Spiegel in Germany around hacking and concerns around Chinese inverters and I think even the EU for recently -- in early May for EU-funded programs ban Chinese inverters. But we've got Intersolar coming up in a few weeks. What are you hearing from sort of either policy at the country level discussing banning Chinese inverters? That would be a big deal.

Yes. So the only thing that was actually decided is the directive that you mentioned that the projects that are funded by European banks, the European banks cannot use Chinese...

They're funded by the governments of the country, not the EU.

Exactly. So that's a very, very small fraction of the market, and people have to realize that. Now the European system, the political system, the way that it works is there are some stuff that -- there is some stuff that is done at the European level and then it's up for the countries to decide. I think that it's a question that each country will have to determine to themselves. On one hand, inverters coming from China are the majority of the inverters in Europe today. So they rely on them. On the other hand, they have some concerns. The sentiment in Europe has some concerns around it. I think that they will make -- whichever decision they're going to make, we're going to continue -- we believe that we can gain share in Europe, by the way, regardless of that. I think that the Nexus that we talked about from a financial perspective also will help us gain share in Europe regardless of the things that you're mentioning.

That's good to hear. Maybe just touching on the battery side of the business. You had the very expensive Samsung cell contract in the past. You're moving to LFP. Is that transition done? And as we move to Nexis and sort of the Clickable Snap, I forget your marketing jargon, but it's a much faster installation time. And for those in the audience that they do a great YouTube series called Night Shift, where literally some guys will sit around over here and talk solar, but if you're ever bored and want to watch.

Not just bored, if you're interested in good customer experience. It's -- Nexis was designed actually in collaboration with many, many installers. And we call it the click, click, click, right? So you come in, you put one module and then you can add additional modules on top of it, which is fantastic. Yes, Nexis is using LFP across the board, which then means that on our single-phase batteries that so far, we've used NMC batteries, we're going to -- NMC cells, we're going to move to LFP. And that by doing that, our entire portfolio is going to be LFP.

Okay. But margins should be better as that transition...

So that's part of the cost advantage that we talked about earlier, right? Nexis is a better cost structure. That's a significant portion of it.

So it's not just the solar piece. And then are these FEOC-compliant cells that will be coming into the U.S. for next year or?

So our battery, the way that it's done is not about the cells themselves, it's about the products that we are making. We're making the product in the U.S. and it's FEOC compliant. It's domestic content compliant, it's everything that our customers need.

And then speaking of FEOC, can we just touch on the C&I business? I think the #1 player in the market, Chint, depending on the day, I think, can be 20%, 40% share depending on the quarter. But what's your expectation of share gains in commercial?

Yes. So you know it, but others might not. In the C&I segment in the U.S., there are 3 leaders: SolarEdge, Chint and SMA. SMA is a German company. And the other 2 companies, they don't have domestic content, and one of them is considered a FEOC company as well. Because of that, the value to the customers that we bring being non-FEOC compliant and domestic content compliant is the 40% tax benefit that they can have. And because of that, what we are seeing is a structural shift of the market share towards SolarEdge. We've gained share in the last 2 or 3 quarters, and we believe that we will continue gaining share, partially because of what I've just mentioned and partially we see that as part of the safe harbor deals that we are signing.

I was going to say, it seems to be interesting if you read between the lines in your marketing message around safe-harboring, you've done a lot of webinars with Novogradac and racking companies for C&I, but less for residential. I guess the TPOs know how to do safe harboring and you don't need to bring awareness. But is maybe something overlooked by investors is the visibility that's shaping up with C&I through safe harbor, because everybody really focuses on...

I don't know whether it's overlooked or not, but we have invested a lot of time and effort and expertise in order to structure a physical work test offering to the C&I customers, which is a different market, to your point, than what the TPOs need. And we feel -- we actually see that this offering is attractive to them, and we are seeing good volumes.

Perfect. Speaking of safe harboring, your rival Enphase was here yesterday, and they had a release out maybe a month ago talking about the cumulative value of physical work test safe harboring, I think, was around $800 million or something, give or take. Have you disclosed that number? Is that something that you would consider doing as you?

So we haven't disclosed the exact numbers, mainly because we are still working on it. So rather focus on the customers, yes, it depends on which time zone. And -- but suffice to say that we're very pleased with the engagement that we've had with customers, both on the C&I side as well as on the resi side, and we feel comfortable with our position.

Perfect. We've got 2 minutes left. Anything from the audience? Otherwise, I had 2 questions on my end. We'll do rapid fire, but -- you've had a herculean effort around improving margins and generating cash. But I think a big chunk of that was the $85 million to $90 million of OpEx target. Do you see with the market in solar rebounding with the SST opportunity, are we at a floor from OpEx? Or is there still efficiencies to be gained as we think about next year? I'm not trying to do a and get guidance for next year.

No. And you know that we don't guide beyond this quarter, right? But I think that the way that we look at expenses, both on the OpEx as well as on the COGS, by the way, we had a concentrated effort on making sure that, on one hand, we reduce our cost. But on the other hand, that we become a more efficient company. The example you gave earlier about single SKU, that will help us save additional cost in the future because we have less SKUs to manage, less SKUs to develop, less SKUs to ship. So we are still actively trying to improve our operational efficiency. And I believe that there is some room to go over there. We have some headwind with the Israeli currency versus the U.S. that does add several million dollars to our OpEx.

Got it. Maybe in the last 30 seconds, but what do you feel is the biggest misperception about SolarEdge among investors? Certainly, you've had a nice run in the stock in the past 2 weeks, but any misperceptions as you've been out there reading?

I wouldn't call it a misperception. I would just say that the way that we look at SolarEdge is we were -- we've been through a rough period of time. 2025 was -- we used it to stabilize the company, to put together all the infrastructure that is needed for profitable growth. This quarter, the midpoint of our guidance is getting there, and we will get there in the future. And so I think that we have a very, very solid solar and storage business that we can build on for the next several years. And on top of that, we're going to add the exciting SSD opportunity. So I'm excited about what's coming.

Perfect. Well, thanks for taking the time out of your busy day.

Thank you. Thank you very much.