Q1 2026 Earnings Call Transcript

Good afternoon, and welcome to First Solar's First Quarter 2026 Earnings Conference Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] Please note that today's call is being recorded. I would now like to turn the conference over to First Solar Investor Relations.

Good afternoon, and thank you for joining us. We're joined today by Mark Widmar, our Chief Executive Officer; and Alex Bradley, our Chief Financial Officer. Mark will provide an overview of our first quarter performance and an update on technology, manufacturing and market conditions. Alex will then cover our bookings, financials and our 2026 outlook.

After our prepared remarks, we'll open the line for questions. Today's discussion contains forward-looking statements. Actual results may differ materially due to risks and uncertainties as described in our earnings press release, other SEC filings and the earnings materials available at investor.firstsolar.com. We undertake no obligation to update these statements due to new information or future events.

We will also reference certain non-GAAP financial measures. Reconciliations to the most directly comparable GAAP measure are in our earnings press release and presentation. This non-GAAP financial information is not intended to be considered in isolation or as a substitute for financial information presented in accordance with U.S. GAAP.

With that, I'll turn it over to Mark.

Thank you, and good afternoon. Beginning on Slide 4. We delivered a strong start to 2026 with record first quarter revenue, record sales in India, meaningful margin expansion, and adjusted EBITDA above the top end of our first quarter preview range. Since our last earnings call on February 24, we secured gross bookings of 1.9 gigawatts, excluding domestic India volume, we booked 1.4 gigawatts into our key U.S. utility-scale market at an ASP of approximately $0.35 per watt, inclusive of applicable adjusters. Turning to Slide 5. Our technology strategy is anchored in the premise that customers value not just nameplate efficiency, but lifetime energy production as well, CuRe is central to that strategy.

Extensive testing data has validated our expected bifaciality, advantaged temperature coefficient and degradation profile. With CuRe anticipated to deliver up to 8% more lifetime specific energy yield than crystalline silicon Topcon. I'm pleased to report that CuRe launch is complete in Perrysburg and the first Series 6 line is ramping consistent with expectations. CuRe is scheduled to be replicated across the Series 6 and 7 fleet through the first half of 2028, which, if achieved, supports the potentialization of up to $0.6 billion of additional revenue from technology adjusters in the backlog with the majority anticipated in 2027 and 2028. Turning to Slide 6. We produced 4.3 gigawatts of modules in the quarter with approximately 3 gigawatts from our U.S. facilities and 1.3 gigawatts from our international fleet.

Our U.S. facilities operated at approximately 96% utilization. South Carolina finishing facility is on track for production start in the second half of 2026, with equipment installation beginning this quarter. Upon completion, this facility is expected to provide finishing capacity for Series 6 modules initiated at our international factories and optimize freight, tariff and domestic content outcomes while benefiting from Section 45X module assembly tax credits. Our international facilities in Malaysia and Vietnam continue to operate at a significantly reduced utilization consistent with current trade dynamics and lower ASP expectations for internationally produced modules.

Turning to Slide 7. Our competitive position in the United States and India continues to strengthen, underpinned by differentiated technology, a domestic manufacturing footprint and bill of material, and independence from Chinese crystalline silicon supply chains. In the United States, headwinds for crystalline silicons continue to build in our view, including trade remedy enforcement, indications of restricted FEOC regulations, and the intellectual property litigation actions we have discussed on recent calls. Our IP specifically in March, the U.S. International Trade Commission instituted our Section 337 investigation with respondents representing a significant share of top 10 modules currently imported into United States.

We expect an initial determination within approximately 11 months and final decision within 15 months. In India, our presence reflects the same strategic logic that underpins our U.S. manufacturing investment, energy security and supply chain independence. The policy framework, including the existing approved list of models and manufacturers or ALMM and the anticipated implementation of the ALMM at the cell level as well as domestic content requirements currently favors vertically manufacturers such as First Solar. Near-term demand is supported by both utility scale and distributed solar applications, including agricultural land developments where our CadTel technologies energy yield in hot humid conditions is a meaningful differentiator.

Overall, our differentiated technology, our domestic manufacturing footprint, and our independence from Chinese supply chains are attributes that are increasingly valued by our customers, and we remain well positioned to deliver on our 2026 commitments. I'll now turn the call over to Alex to discuss our bookings, financial results and outlook.

Thanks, Mark. Beginning on Slide 8. As of March 31, 2026, our contracted backlog was 47.9 gigawatts at an aggregate transaction price of $14.4 billion, exclusive of technology adjusters with deliveries through 2030. During the first quarter, we sold approximately 3.8 gigawatts, reported gross bookings of approximately 1.7 gigawatts and recorded debookings of 0.1 gigawatts. In India, our guidance assumes that production is largely sold domestically in a book-and-bill market at near full capacity. In the first quarter, we sold approximately 1 gigawatt in country at an average selling price of approximately $0.20 per watt.

In the United States, our domestic production has substantially committed through 2028 under existing contracts, resulting in relative pricing clarity through this period. We continue to take a highly selective approach to incremental U.S. bookings as we await the outcomes from current policy and regulatory matters, in particular, the pending 232 polysilicon derivatives tariff decision and proposed FEOC rulemaking. During the first quarter, our gross bookings of 0.9 gigawatts were at an average selling price of approximately $0.34 per watt inclusive of applicable adjusters. With respect to our international fleet, demand for Series 6 modules produced end-to-end in Malaysia and Vietnam remains constrained, which is reflected in the reduced production Mark mentioned earlier.

Turning to Slide 9. Net sales of $1 billion were a record first quarter for the company and grew 24% year-over-year. This is driven by a 31% increase in volume, partially offset by lower average sales price reflecting a higher proportion of India deliveries. Our gross margin in the first quarter was 47% and expanded approximately 6 percentage points as compared to the first quarter of 2025. The drivers were primarily a higher volume of modules qualifying for Section 45X tax benefits and significantly lower sales freight costs, including lower detention and demurrage. On a per-watt basis, sales freight costs for approximately half of first quarter costs last year at approximately $0.017 per watt.

Furthermore, as part of our plan to rationalize warehouse costs to approximately $100 million by 2027, we delivered a $22 million sequential reduction in warehouse costs from Q4 2025. On balance, these savings were partially offset by a lower average sales price due to a higher mix of India sales and increase in tariff costs year-over-year. Operating expenses for the quarter were $141 million, including R&D of $67 million, up $15 million year-over-year, primarily reflecting perovskite development and ongoing CuRe launch work. Adjusted EBITDA was $520 million, above the high end of our first quarter preview range of $400 million to $500 million.

Adjusted EBITDA margin was 50%. Net income was $347 million, up 65% year-over-year with diluted EPS of $3.22. Moving to Slide 10. We ended the quarter with $2.4 billion of cash, cash equivalents, restricted cash and marketable securities and a net cash position of $2 billion, at the high end of our targeted resilient net cash range of approximately $1.5 billion to $2 billion. Operating cash outflows of $215 million reflected normal first quarter working capital dynamics, a meaningful decrease from outflow of $608 million in the first quarter of 2025. Capital expenditures were $119 million, primarily for our South Carolina finishing facility.

We also completed a $45 million scheduled principal payment on our India DFC loan. Turning to Slide 11. Our full year 2026 guidance remains unchanged. For the second quarter, we expect volumes sold between 3.4 and 4 gigawatts and adjusted EBITDA of $400 million to $500 million. In summary, our first quarter performance and reaffirmed guidance for the full year reflect the strength of our strategy of reshoring and scaling domestic manufacturing, progressing our technology road map, enforcing our intellectual property and maintaining a selective approach to new bookings in light of key pending trade and policy determinations.

And with that, we conclude our prepared remarks and open the call for questions. Operator?

[Operator Instructions] Your first question comes from the line of Brian Lee from Goldman Sachs.

Just had 2 here. First, on the module gross margins, I think it's around 7% in Q1, when we adjust out the 45x even with the high India mix. So curious, the guide for Q2 implies margins are flattish. Why not more improvement given some of the sequential improvement in freight costs, warehousing, et cetera?

And then what kind of tailwinds maybe help into 3Q and 4Q for the margins beyond just volume growth? And then on the ASPs, maybe I'll just squeeze in the second question here. This could be nitpicking, but on Slide 8, you show 900 megawatts of U.S. bookings, $0.34 in Q1. You mentioned 1.4 gigawatts since the last quarter at $0.35. So it maybe implies recent bookings, March and April are higher at $0.36, $0.37 per watt. Anything to read into that?

What kind of customer engagements and discussions are you having here more recently ahead of more policy certainty?

I'll do the ASP first, Brian, and then Alex will talk through the gross margin. ASP-wise, yes, so the 1.4 is call to call. So that's a little bit of maybe a clarification there. So -- which would include a half and half.

So half of that happened before the quarter end. All that happened after the Fed call, but half of it happened before the quarter end, the other half happened after quarter end. And the average ASP for the call-to-call volume of that 1.4 was $0.35. One other note to include in there is that there's about another 700 or so that's an option. We're seeing a lot of M&A activity.

We're seeing a lot of -- look, what's great about First Solar, we've had very strong strategic partnerships with very confident obviously well-capitalized partners, and they're actually seeing a lot of development acquisition opportunities. Actually talking with the team about another deal right now that one of our customers is actually in the process of acquiring is looking for incremental volume to support that acquisition. So what happened was we booked a deal for about 700 megawatts. Our customer is actually now in the process of looking to acquire another development asset of which then they would exercise that option, which would have to be exercised here over the next several quarters upon completion of that acquisition.

So what I would just say, there's still a lot of momentum, a lot of activity going on, Brian, from a market standpoint. We're being very disciplined as we have in the prior quarters with how we're engaging the market, and how we're seeing through pricing. So still good momentum, discipline on our part, trying to realize good ASPs, which I think we did here in the last, call it, 8 weeks since the last earnings call.

Brian, on the gross margin, so if you think about India on an aggregate cents per watt or dollar contribution basis, it's certainly lower than the U.S. If you look at where India is pricing today on a gross margin percent basis, it's not materially different to the U.S. So despite having a higher India mix on the percentage gross margin basis, not a material impact. If you look at the full year guide of 7%, we haven't changed the guide, that still holds.

That's right where we came in ex-IRA benefit in Q1. As you think through going forward, next quarter, we're guiding to the same guide that we had in Q1. So it's relatively flat, which implies the second half is going to be stronger. Incremental volume should be beneficial to gross margin. There's a little bit of value on the fixed cost side. The other piece in terms of the back end of the year is that right now, our assumption on tariffs is that the Section 122 tariffs carried through for the 150 days from announcement takes us through to the July timeframe.

After that, we're not modeling tariffs beyond that for finished goods coming in, right? There's still other tariff impact in there on 232. Now that's the modeling assumption we have today given uncertainty around what could happen. As you're probably aware, the Trump administration has launched several 301 cases with the intent, we believe, to try and replace the 122 with that 301 later in the year. But as of now, we're not modeling that.

So I think the guide has stayed where it is, the guide assumes our tariff replacement back end of the year, that would imply you might get some incremental gross margin. However, if we do see additional tariffs, we obviously reflect that in the guide later on.

And I think maybe one of the things, too, just around operationally quarter-on-quarter from a gross margin standpoint, we will be running -- we ran Malaysia and Vietnam at higher utilization rates in the first quarter than we anticipate running in the second quarter. So you're going to see more underutilization charges in the second quarter. So trying to get to your question, Brian, about sequential gross margin performance, we will see a little bit of headwind in the second quarter because of the lower utilization rates for Malaysia and Vietnam.

Your next question comes from the line of Julien Dumoulin-Smith from Jefferies.

This is Dushyant for Julien. I guess two quick ones. Could you quantify the amount of adders, I guess, on that $0.34 or $0.35 ASP? Is it like $0.02 to $0.03 that you discussed before? And then second, a real quick one, how do you think about the Southeast Asian capacity going forward?

I'll take that. I'll do the adders and Alex, you can talk to the Southeast Asia thought process and where we are on that. So one of the things is now with the launch of CuRe as an example, we will -- you'll see it's kind of -- I guess, what I want to make sure people understand is the product at which we're going to price now going forward, given that CuRe product is being launched is going to be the technology which we anticipate to deliver in this full entitlement. And if you look at the bookings that we reported, call at 1.4 gigawatts since the last earnings call, half of that volume actually sits out in 2029. So it's encouraging, we're starting to see more momentum in the outer years.

And that's a CuRe product that we expect to deliver. And so when you look at the adder and that in some cases, the volumes that we're seeing right now, there are no adders because we're pricing the technology. Therefore, you won't see an adder to that deal. And so when you look at the blended average of the adder, the entitlement of the adders are still, call it, $0.03 or so, but half of the volume that we booked did not have adders on it, the other half did.

So the blended average of the adder is going to be call it, $0.015 or something along those lines. But I think what we'd like to be able to do as we move forward, especially now with the launch of CuRe. We're pricing the technology. We're going to deliver the technology, we're going to price it.

All the energy attributes will be embedded in the base price. We're kind of still in that transitionary period. We're going to see some contracts that potentially have the combination of 2, the base being our current semiconductor, which we refer to as QD plus adder, but for the windows, in which we for sure are going to be delivering the CuRe product, we'll just price it as the contract, and you won't necessarily continue to see the adders. I just try to make that clarification.

So as we transition in that direction, you're probably going to see less volume with adders on and just full entitlement of the technology being priced.

As it relates to Malaysia, Vietnam, we talked on our previous call about maintaining an option around that capacity, and we're still doing that. If you look through what we're waiting. I think a large part of that is policy clarity around the 232, which really could spur potential demand around the fully finished international product. If you go back to the original guide, we had $115 million to $155 million of underutilization costs.

That was a function both of Malaysia, Vietnam underutilization running at very low capacity through the year as well as some of those costs associated with the new finishing line that we're bringing up. So right now, we're continuing to maintain that near-term option. We'll continue to evaluate that. Likely, the decision point we're waiting on is policy clarity around the 232, which we expect most likely to come in Q2 of this year.

Your next question comes from the line of Christine Cho from Barclays.

I actually wanted to know if we should think that there's an impact from the Section 232 changes on steel and aluminum for the Southeast Asian imports. I was under the impression that aluminum might be less than 15% of the weight of a full module, but now that you're going to be just importing just the front of the module, like how should we think about that? And then I think you also said on the last call that 5 gigawatts of the backlog was international modules. Was the plan just to mostly import that at the beginning of the year, and it tails off at the end of the year?

If you could just talk about the cadence of that as well.

So I got the aluminum. Repeat your second question, I want to make sure I have it.

My second question, the 5 gigawatt of backlog was -- I thought that was the international portion. Was it not?

Yes, yes. No, that's about right. I think that's approximately the number. So yes, so that was entering the year, we had about 5 gigawatts of S6 international backlog. And so from there, your question was?

Was the plan to mostly import that at the beginning of this year?

No, no. So okay, I got it now. Sorry, that's a piece I didn't have. Yes, that 5 gigawatt of backlog on S6 international was a multiyear backlog.

Those shipments would go across '26, '27 and into '28. So that was a multiyear. So we will run that production to meet that demand of that portion of the backlog, call it a little bit more than 25%, around 25%. It would be for this year. The balance would sit in the outer years as it relates to that.

On the 232, yes, aluminum is still included in the tariffs for 232. I think part of your question was as well, what happens with the semi-finished product that comes into the U.S. That will come in -- the glass will come in, obviously, without the aluminum frame. As of now, we will continue to import the aluminum frames into the U.S. And because they continue to be imported, they'll be subject to the applicable rates associated with the 232 for aluminum.

So anyway, yes, 232 is still applicable for aluminum based on the classification of the product we're bringing in, and no real change to that just because we're bringing in a semi-finished product from Malaysia.

Okay. And then you guys have -- your -- you guys have been doing a lot of India volumes in this quarter and the full year assumes full utilization. I just saw that there was like in one of your disclosures about India's, there's a proposal in India to increase the minimum efficiency of PV modules for manufacturers to be included in the ALMM beginning in 2027. Could you talk through exactly what's going on there and what your options are to work around it? I realize that it's still just a proposal, but?

Yes. A couple of things. There's a lot of moving things that are going on in India, even including the requirements by 2028 to have a qualification of the wafer being domestically manufactured in India in order to qualify for any projects that actually connect to the grid, the federal grid or the state grid. So that's moving is one of the moving pieces that will, I think, further enhance our opportunity to continue to support the Indian market given our vertically integrated model.

The other one as it relates to the consideration for the efficiency threshold. We will be launching CuRe beginning of next year in India. So our first Series 6 facility that we'll launch in India will be CuRe, which will give us an opportunity to improve the efficiency of the technology as well as continuing to enhance the energy attributes, better bifaciality, better temperature coefficient, better long-term degradation rate. And as we continue to work with MNRE, in particular, in other parts of the administration and continuing to inform them and educate them on the value of the attributes of energy, which is also what our customers pay for energy, not labeled efficiency, we continue to have very good and constructive dialogue in that regard.

But I think the key enabler to make sure we manage through any potential revisions to those requirements will be the launch of our next-generation technology of CuRe in India.

Your next question comes from the line of Chris Dendrinos from RBC Capital Markets.

I just wanted to follow up on the earlier question from Jefferies in regards to the Southeast Asia offtake agreement. Can you maybe walk through a bit of the possible decision tree here, depending on, I guess, the tariff outcome or an offtake agreement, and how you're thinking about that facility? And when would we have a potential decision as to what you all do longer term?

Yes. I mean right now, there's ample demand for the product at a certain price, but when you factor in the tariff implications of bringing that product in and then the risk allocation around that tariff, it's not necessarily the right risk profile for us to take today. Depending on an outcome of the 232, you could potentially see much higher pricing or risk tolerance from buyers whereby we will be able to more appropriately price that product. And more appropriately allocate the risk around changes in tariff policy, which will then enable us to be comfortable long-term contracting that product.

So a lot of it depends on availability of supply in the U.S. and where tariff and risk can be allocated. And so we think the 232 is most likely the main determinant of that. The outcomes there could be, we continue to run that product at full capacity end to end and ship fully finished goods into the U.S. it could be that there's demand, we could add an incremental finishing line in the U.S. and finish that capacity here? Or it could be that neither occur and then we're into the potential shutdown of that capacity.

So those are really the pieces we're looking at.

Yes. And one thing maybe just to help -- remind everyone of what we showed in the last earnings call. So historically, we've had about 7 gigawatts of capacity between Malaysia and Vietnam, okay? Half of that is now going to come to the U.S. to support our South Carolina finishing line.

So 3.5 of that, that leaves the other 3.5. Now what we indicated in the last earnings call, about half of that 3.5 is largely no longer available because we're moving some of that back-end capacity to the U.S. to support our perovskite pilot line. So perovskite full-size Series 6 pilot line will be available in 2027. So we lose the back-end capacity. As a result of that, we're losing some of that throughput for the facility. So when you really look at how much of the Malaysia facility of that 3.5 gigawatts that would be available, which will be fully manufactured, assembled modules to be shipped into the U.S., there's only about, call it, 1.8 gigawatts, maybe closer to 2 of real capacity.

So what we're talking about is from a full-size finishing module capacity perspective, there's slightly less than 2 gigawatts that is in play that would tether back to a 232 decision point. So I just want to put that in perspective. As you think through your analysis and assessment of Malaysia, Vietnam half of it's already going to be coming to the U.S. as a semi-finished product. Half of it has already been -- capacity has been reduced because we're moving the back-end tools to help support our perovskite pilot line.

And then now we're kind of still with about slightly less than 2 gigawatts in Malaysia, Vietnam, but it will be tethered back to whatever decision is made with 232.

Your next question comes from the line of Phil Shen from ROTH Capital Partners.

On the 232, I was wondering if you might be able to share a little bit more color on what the framework of that decision might be. I think we published earlier this week that there could be a minimum import price in the $0.38 per watt kind of level. And so does that resonate with you guys at all? And on your Slide #7, you talked about the timing being Q2. We've seen this push a bunch, right?

It was supposed to be year-end '25, but then the shutdown happened, and a bunch of other reasons have driven this a little bit later. And so as we are a month into Q2, what's the confidence level that it comes out in May or June? And then on the technology front, Mark, I think you just talked about a full-scale line of perovskite in 2027, which is really interesting. And so I was wondering if you could give us a little bit more color on that?

What kind of costs are you seeing? And then is that -- what's the base? I got to believe it's CDT on the bottom cell and then perovskite on the top cell. But just any other color in terms of efficiency, durability, et cetera?

Yes. So Phil, on the 232 and the framework, I mean, look, there's still a lot of moving pieces, right? What all I can say right now is that the engagements that we're having with the administration and the structure of which we are trying to propose, which again is not 1%. It's basically $0.01 per watt on the cell or $0.01 per watt on the module, or could include a minimum import price to your point.

The feedback we continue to get is very positive to looking at that as the best way to address the polysilicon and its associated derivatives. So it's a good positive feedback. But Phil, as you know how these things play out, they continue to evolve. And so we have to stay as well connected as possible to continue to advocate, advocate and advocate for that type of position.

But what I would say is, at least, is it's encouraging. As it relates to the timing, similar the feedback that we are getting is still a resolution by the end of this quarter, does it move into early Q1 potentially? But it's all dependent on other events that happen, right? There's always -- I think the intention is that they know they need to bring this to a conclusion.

It's important and they need to communicate an outcome. But as you know, Phil, these things can always move. So is there a level of certainty for the quarter? What we just represented in terms of the slide is the information, the best information we have and what our expectations are based on those communications around an outcome and communication around 232. On the technology side, yes, so we are currently in a time line that would have us running a pilot line in 2027 for perovskite.

The cost piece of it, I'm not going to get into the specifics of that. But that pipeline, which is a 1 gigawatt pilot line, by definition, it's not going to be an HVM type of cost entitlement, right? Because in order to get cost out, you need to run high throughput and scale, right? That's like if you go back and if you -- if you look at when we went from our Series 6 factories to our Series 7 factories, they're -- we tripled the output effectively, and we managed it with the same headcount.

So no real dramatic change to headcount, and that helped drive cost out significantly. And so at 1 gigawatt, you're going to be suboptimized, and cost is going to be higher, but for an initial product to get it into the market and to get field validation and customer feedback and all the other things that we need, we think it's appropriate to do that upon launch. And then as we continue to evaluate, we'll move into a scaling mode. But again, it's going to be a higher cost product.

There's no doubt about it upon launch. As it relates to the construct of what the product is? And is it a single junction, is it tandem? I mean there's -- we're looking at 2 different paths.

We're still evaluating what we believe is the right launch product. I think the most important thing to do to get something in the field is to validate the performance out of the perovskite. If you can do that and reduce some complexity of thinking through the additional challenges of integration of the tandem construct because you have 2 different products that have different electrical properties, right? Different temperature coefficient as an example, and different electrical properties, voltage current and go on.

To try to deal with the construct of matching those 2 different semiconductors and optimizing them and harmonize them at the module level. So they perform as effectively a system if you want to refer to it as such. There's a level of complexity around that, that I'm not sure it's worth the effort initially because what you want to do is get the product in the field, validate its performance, validate its degradation, validate its performance in all forms of conditions, right, which would include open circuit, how does it perform in partial shading. So there's a lot you can learn.

And I think the most important thing you want to learn is that the perovskite level. We all know how -- whether the tandem construct would be a perovskite with a CadTel module or perovskite with a silicon bottom cell. Those you can learn and evolve, but what you really need to do is validate the durability and viability of the perovskite, and it can perform in the field, and be a bankable product and can hit the long-term performance objective that it's going to be held to.

Your next question comes from the line of Vikram Bagri from Citibank.

My first question, Mark, probably for you, We understand the market is for contracting panels at $0.33 a watt or higher is not as the customers are hesitant in pricing the Section 232 risk as of now. I was wondering like once it comes out based on your assumption, sometime in late 2Q. How quickly can you move to book volumes? Put it another way, how much demand is beating for Section 232 to come out?

What's sort of like what should we see or look for in bookings after immediately after the Section 232 comes out?

Well, look, I think there is some unknowns in there as well. I mean, once it gets communicated, then the other is to what extent is -- what is the impact, right? And that will be a key component of how much volume and how quickly. I mean, what I will say is that we have some customers that are looking at multiple gigawatts of volume and they're waiting.

And what we've also said to some of those customers that once this gets communicated, depending on what the outcome of that communication is it could have an impact on the current price that we're at least negotiating. So the risk we run is that it ends up being lower than anticipated, the risk they run, and it ends up being higher than anticipated. So I think there's quite a bit of demand there that should provide an opportunity for us to move through and to book that over a multi-month period of time. But it really depends on the outcome, and that's really what the -- where you got to bid ask right now.

That's what it relates to is what do we expect that outcome? We're trying to create a price point, we think that's reflective of the midpoint, and we'll see what comes out. If we're wrong, then we may see a little bit of ASP pressure. And if we're right, we may see a little bit of upside relative to that marker we have engaged the market with right now.

Your next question comes from the line of Colin Rusch from Oppenheimer.

It's a 2-part question. So first, as you move from Series 6 to Series 7 with CuRe, can you talk about the key technical elements that you guys are working on right now and things that we should be watching for, for success? And then can you talk about any sort of impact that you're seeing from the incremental capacity, domestic capacity that's coming online to some of your pricing negotiations?

Yes, Look, the -- from a technical standpoint, Series 6 to Series 7 isn't really a significant technical challenge. What it is, though, is -- because it's just a form factor change for the most part, right? We are -- for the Series 7 launch -- so right now, we are for the front contact buffer, we actually are taking our existing TCO glass, and then we're actually depositing the front contact buffer for Series 6. But for Series 7, our plan would be because we've been working with our glass suppliers to allow them just to do it within their facility. So we no longer we'll just get our standard TCO.

We'll get the glass will also be included -- include the revised front contact buffer that we need for CuRe. So that will simplify the factory operations from that standpoint. But really, it's just the difference of the size of the tools. So the tools that -- there are a couple of new tools that we've had to add, it's because of the BKM, best-known method and process is different for CuRe and our existing product.

And so there are a couple of new tools. Again, those are first-of-the-kind tools that have been designed and spec for a smaller form factor product in Series 6. So we have to now operate those tools, season those tools to get them to the performance level that we need for a different size form factors. So it's not as much of a technical challenge. It's just that there's a slightly different process that we're using because in Series 6 and at least the launch, we're doing the front contact buffer in Series 7, we won't be doing it, and then the difference in size of the tools, is really what the real challenge is.

So which -- we're working through that. We're validating all that right now. And we're going to go as fast as we can. Once we get comfortable with validation of a lot of that, we'll move forward and try to replicate as quickly as we can across the fleet.

The new capacity, not -- so if you exclude our fully integrated capacity between Ohio, Alabama and Louisiana, the new facility in South Carolina is going to be semi-finished product. So what's nice about it is it's a Series 6 form factor. It does have domestic content. It doesn't have as much domestic content is a product that we manufacture in Ohio, but it creates a nice opportunity to blend the two together for our customers.

They get value out of that as well. It does sort of create an opportunity to be a little bit more competitive on pricing. It creates an opportunity to create more domestic content value to our customers, and therefore, enabling a broader portfolio of projects that can benefit from the domestic content bonus, which is extremely valuable to our customers, especially on the ITC side, which in some cases, we're talking at the project level, $0.15 or $0.20 or more of value to our customers to enable that bonus. And so it's a nice way to sort of enhance and further create value to our customers' portfolio by enabling them to see higher realization against domestic content bonus for ITC in particular.

Your next question comes from the line of Praneeth Satish from Wells Fargo.

With the IPA tariffs repealed and import tariffs on India-produced modules down to 15%, just seeing how are you thinking about selling the India capacity into the U.S. versus selling it in the Indian market? Is that something that you're still considering? Maybe if we get a positive 232 outcome, and then to the extent that you are, what's the lead time and retooling costs that would be required to enable that? And then just a quick housekeeping question on the 1.7 gigawatts of bookings this quarter.

Are you able to break out roughly how much of that is from U.S. capacity versus international?

The -- on the India -- look, I mean, right now, there's a lot of demand in India. We just sold 1 gigawatt last quarter. And when you look at the actual gross margin on that product effectively the highest gross margin that we have, excluding the benefits of 45X. So the gross margin is on a percentage basis is obviously attractive.

The -- and the changeover is always -- downtime -- even though we try to optimize to make it more efficient to change from a fixed tilt product to a tracker product. It still is not efficient to do that. So we are right now looking at this from a lens, of let's just keep running that product because of the demand that we're seeing, at least through the first half of the year. Q3, we see a little bit of softness generally in India, and then you'll see a stronger Q4. So there maybe a little bit of volume in Q3, assuming the tariff environment stays where it is, the challenge with the 122 expire at the end of July.

So we're going to be looking at whatever tariff environment would be enduring after that. I don't -- we have no idea what will happen with the 301s. And so it's kind of a watch card. We'll see what happens.

We will be bringing some volume over, we're talking 10s to low 100 megawatts or so of WIP share product from India and bringing that into the U.S. and finishing here in the U.S. We're doing some of that in the first half of the year to help enable demand here in the U.S. market. But yes, it's something we continue to evaluate. I just think we just don't want to -- until we have a better understanding of what the long-term tariff environment is going to be because, again, we don't know after the 150-day window for the 122 is what exactly a tariff environment will be in.

I think for now, we're going to focus on a market that has very strong demand continue to support it from that standpoint.

Praneeth, your housekeeping question, of the 1.7, 0.9 was U.S. and 0.8 was India bookings.

Your next question comes from the line of Maheep Mandloi from Mizuho.

Just maybe some housekeeping on the first half versus second half cadence, if I look at the EBITDA versus the volumes, it looks like the EBITDA is 36% in the first half, but 44% of the volumes. So just trying to understand, is that because of India shipping is higher in the first half? Or is this SG&A, which is kind of skewing the EBITDA in the first half or second half?

Yes, it's mostly driven by the India volumes. As Mark said, we had a strong Q1 for India. It will drop down in Q2, Q3 and potentially pick back up in Q4. But as of now, the guide assumes that imbalance, which is why you're seeing that.

Your next question comes from the line of Joseph Osha from Guggenheim Securities.

I'm still trying to understand the composition of this 3.5 gigawatts in South Carolina. Are you saying that we have for starters with perovskite part of that? And then you're saying that absolutely no matter what happens, we're going to run 1 gigawatt and changes to it through S6, and the rest is optional. I'm trying to put together where this 3.5 comes from.

Yes. Thanks for that question, Joseph. That 3.5 in South Carolina is our CuRe or will be our CuRe product and Series 6 product has nothing to do with perovskite. So it's going to be a product that will be started, think about it just basically the substrate glass with the deposition on it with cell scribing, and then shipped to the U.S. to be finished, which will include the cover glass junction back -- junction boxes, the frame, in layer, all the other stuff that's going to be added to it, right?

So that's what that product is, right? So it is a CadTel product, okay? In addition to that, we will be launching a pilot line next year, which we communicated this as well on our last earnings call, we made the announcement that we acquired the IP for Oxford for perovskite, which enhances the IP that we already have to manufacture that product here in the U.S., and we will be starting that with the pilot line that will have up to 1 gigawatt of capacity. That pilot line will actually be in our Perrysburg facility.

So we have existing facility space in Perrysburg that we will be integrating and using for that pilot line, and we'll be leveraging some of the capabilities that we already have from a back-end processing and what have you in Perrysburg. So that's what that means, right? So it's a pilot line. We still think of it as development has 1 gigawatt of capacity.

We will be manufacturing product when we deploy it into the market, getting test data validation with customers, and those types of things. South Carolina will be a CadTel product, started in Vietnam, Malaysia, finished in South Carolina.

And maybe the part of confusion there. So if you think about the original 7 gigawatts of capacity we had in Malaysia, Vietnam, 3.5 is still going to be used for the front end of the product that then comes to be finished in South Carolina. Of the remaining 3.5, we're using some of those tools. Those tools will then go into the perovskite line in Perrysburg.

So that's why the remaining CadTel capacity in Southeast Asia comes down, but it's not mixing CadTel and perovskite and simply saying some of those back-end tools will now no longer be used there. They're going to go over to Perrysburg and going to be used in that 1 gigawatt pipeline for perovskite.

Our final question comes from the line of Ben Kallo from Baird.

I just have a follow-up question, and then another one. Just to Joe's question, is -- what is the capacity after all of that's done? Because I think, Mark, you said that you lose some capacity in Vietnam and Malaysia. And so I just want to make sure that we have the volume number correct as we enter next year.

And then my follow-up question is just on TOPCon and your patent and what Tesla is doing, and how you think about them starting manufacturing here and if that's going to violate your patent.

Yes, Ben, if you go back to the deck that we presented in February, we gave capacity and production for '26 and '27, and the assumptions at that point have not changed to now. So if you look at on a production basis, we said for 2027 will be around about 19 to 20.5 gigawatts. So 19.7 I think, was the midpoint of production, and that's not changed. So if you go back to that deck, you'll see the breakout there in terms of geographical location of product.

Yes. And Ben, on the Tesla as it relates to our TOPCon patent, what we do know is that any of the TOPCon product, as you can see by the filings that we've done, and the number of manufacturers who have produced TOPCon and have sold it into the U.S., effectively, all of those parties have been infringing on our IP. Okay? If Tesla chooses to go with TOPCon, my assessment would be given what I see in the market unless Tesla tries to redesign the product such that they would not infringe our IP.

I guess my assumption would be there will be some form of infringement, okay? And look, I also want to make sure that it's clear that we are more than willing to work with any counterparty to engage in a commercial conversation around the licensing of our IP. We are not prohibiting that conversation, right? The issue is, we just want to be paid fair value.

That's also why we license the IP to talent. The talent has demonstrated a willingness to pay fair value for the technology that's enabling the product they're going to manufacture. That's fine. And we'll do that with other counterparties.

And if Tesla chooses to use an IP TOPCon product, it uses our IP, then we'll center to a commercial conversation with them and happily engage with the conversation of licensing that IP. I mean, there's nothing wrong with that, and we'd be more than happy to do that. I do think it's one of the challenges that Tesla is going to have to figure out what technology they go with, and how do they get freed up to operate. And IP is going to be a significant challenge, at least in my mind, as it relates to TOPCon because it's a challenge that everyone else here in the U.S., I think, has realized as well, particularly as it relates to the strength of our IP, but more I'm more than happy to enter into a commercial conversation with them, if they would choose to do that or want to do that, and we'll figure out a license arrangement that can work for both parties.

Look, I think Tesla getting -- establishing capability here in the U.S. market. What we've always said is we need a robust and resilient domestic supply chain completely vertically integrated beyond just thin film CadTel, that's also why we're evolving towards perovskite as next-generation thin films. Tesla bringing in that capacity and that capability and creating a domestic supply chain. I think it only further enhances and supports kind of the overall strategic intent around long-term energy independence and national security and having domestic supply chains, I think, are extremely valuable in both of those regards.

At this time, there are no further questions. This concludes today's call. Thank you all for attending. You may now disconnect.