Home / Transcripts / Tower Semiconductor Ltd. (TSEM) · November 2, 2026

Tower Semiconductor Ltd. (TSEM) Earnings Call Transcript

November 2, 2026

NASDAQ US Information Technology Semiconductors and Semiconductor Equipment earnings 67 min

Earnings Call Speaker Segments

Noit Levi-Karoubi executive
#1

Hello, and thank you for being with us. Welcome to the conference call on Tower Semiconductor's financial results for the third quarter of 2026. Joining us today are Mr. Russell Ellwanger, Chief Executive Officer; Dr. Marco Racanelli, President; and Mr. Oren Shirazi, Chief Financial Officer. Before we get started, please be aware that some statements on today's call may be forward-looking and carry risks and uncertainties that might cause actual results to differ materially. Those risks are described in our SEC filings, Form 20-F and 6-K, along with filings with the Israeli Securities Authority, all of which you can find on our website. Tower undertakes no obligation to update forward-looking statements. Our results for the third quarter of 2026 are prepared in accordance with U.S. GAAP. Certain data we present may include non-GAAP financial measures as defined under SEC Regulation G. Reconciliations to GAAP figures and full explanations appear in today's press release and financial tables. For your reference, a supporting slide deck is posted on our website and integrated into this webcast. With that, I'll hand the call over to our CEO, Mr. Russell Ellwanger. Russell?

Russell Ellwanger executive
#2

Thanks, Levy. Hi everyone, and I appreciate you joining this call. We have plenty to discuss. We'll talk about how impressive our present results are, and — crucially — what's coming next: specifically, the enormous operational and monetary strength we'll carry into 2027, how clearly we can see upcoming demand, the payoff from investments designed to seize all of that demand, further client commitments covering 2028 and 2029, and our considerable expansion in value-creation efforts aligned with customers for 2029 and afterward. This really is a long-horizon value-building push with people on the ground, producing financial and strategic wins you can measure as they happen. During Q3, we continued posting record results while strengthening our technology leadership and adding capacity. Growing customer commitments are underpinning a much higher level of revenue and earnings in the coming quarters and years. Our silicon photonics platforms are central to this growth, as surging AI infrastructure demand, expanding customer programs and long-term capacity commitments keep scaling Tower's operations and strategic position. Alongside this momentum, we are executing well across our wider portfolio of differentiated analog semiconductor technologies, and we are entering the strongest growth phase in our history. The third quarter's performance demonstrates that our strategic choices are producing record revenue, greater earnings power and substantial visibility into the years ahead. For the third quarter, we brought in $526 million in revenue, up from $396 million during the same period in 2025 — that's 33% growth year over year. Net profit came to $120 million, a 23% net margin, compared with 14% a year earlier. Looking ahead to Q4 2026, our mid-range revenue guidance sits at $600 million; hitting that figure would put us at $2 billion for the full year 2026, versus $1.57 billion for all of 2025. Reaching a $2.4 billion revenue run rate matters a great deal, since our demand picture and start plans don't show the usual first-quarter seasonality the industry tends to see. Quite the opposite — we anticipate further growth. So because of that, we're aiming for the 2027 starting point to be above the $2.4 billion annual begin run rate, and more than a 45% increase relative to where first quarter of 2026 revenue starts. Later on in this call, we'll give you details that highlight the significant quarter-over-quarter growth we're generating in 2027. Now turning to our businesses. As a reference for the Q3 revenue breakdown by technology, please see Slide 4. In the third quarter, our teams once again delivered strong operational execution, both in expanding and repurposing 4 of our 5 factories toward a substantially higher infrastructure mix, with growth activities underway across all of our core businesses. Silicon photonics is growing extraordinarily well for us. We're adding capacity at four manufacturing locations to keep pace with rising customer demand, and that business generated about $240 million in silicon photonics revenue in Q3 2026 by itself. A substantial amount of that capacity comes online this quarter, so we expect year-end capacity to be close to double what was in place to support the Q3 2026 silicon photonics revenue. This positions us for a substantial revenue increase across Q1 and Q2 of 2027, as those new start rates move through the fab and ship out to customers, establishing a much stronger sustainable baseline for subsequent quarters. Beyond that, we will also recover our full ROI inside one year on the $920 million in CapEx that we previously announced. In addition, our released Track-1 300-millimeter expansion is executing ahead of plan — all tools are ordered and slotted, and we are targeting the start of full production in Q3 2027. If you factor in the Track-1 300-millimeter expansion, we're aiming to grow our silicon photonics starting capacity by three times what it was at the level that backed Q3 2026's $250 million in SiPho — $240 million in SiPho shipments. The key point is that this capacity is already claimed. By combining signed agreements with contracts that are in advanced negotiation stages, we've locked in customer demand commitments spanning both 200-millimeter and 300-millimeter production, and we've now stretched that coverage out to 2029, a point at which our incremental Track-2 plant capacity ramp is well underway and will ultimately quadruple our 300-millimeter capacity in Japan. One thing to note: all of this expansion is intended to stay self-funded. In addition to addressing the fast-expanding segment—mostly pluggable scale-out transceivers—we keep up numerous initiatives with Tier 1 customers that capitalize on fresh possibilities widening the market we serve. Chief among them is the move in form factor toward near package optics, or NPO, which is expected to bring scale-up prospects on top of the leading position we have clearly built in scale-out; according to most analysts, NPO scale-out could make the optical transceiver market twice as large. With that goal in mind, we are building NPO prototypes alongside over 10 of our biggest customers. With NPO, we're frequently able to put both transmit and receive capabilities onto our SiPho die, which widens our silicon SAM compared with the standard pluggable market, where typically the SiPho die carries only the transmit side. NPO also stands to strengthen our standing in the electrical IC SiGe market, since most NPO implementations will swap out the DSP for SiGe TIAs and drivers to bring down cost, power and latency — and those parts will occupy more footprint as equalizer functionality gets incorporated further, which broadens our SiGe market opportunity. Along these lines, during the quarter we announced a significant capacity agreement with Semtech, and we're seeing even more unprecedented growth opportunities for our SiGe business. Looking further out, we're developing the ability to bring new materials onto our SiPho platform, which would widen the served available market even more. As one illustration, we announced initial production of new photonics built with a new technology in which indium phosphide lasers are integrated inside our silicon photonics die. That's achieved by attaching indium phosphide InP chiplets die-to-wafer, which then get further processed in our factory to produce a final die that incorporates indium phosphide lasers within our silicon photonics. As this technology scales up, we unlock entry into the laser market along with extra multiples of growth in our serviceable market. We're in the process of deploying this die-to-wafer method with more materials, such as thin-film lithium niobate, so that next-generation differentiated products at 400 gigabit per lane can reach our Tier 1 lead customers. Beyond optical transceivers, our SiPho platform combined with heterogeneous integration has emerged as a preferred choice for optical circuit switches built on silicon photonics. Solid-state OCS based on silicon photonics is gaining ground as a serious alternative to MEMS and other specialized approaches, because of nanosecond-level switching speeds, CMOS manufacturing scale, and compatibility with the CPO/NPO ecosystem that enables tighter system integration. Finally, as we've noted before, we've been a major supplier of coherent SiPho products for Marvell in data center connect applications. Teams are currently assessing this technology for deployment at data-center scale across various applications, and that could greatly broaden the SiPho market segment as well, a space where we already have a solid foothold and distinctive technology. Looking further out, we're putting investment into multiple scale-up and scale-in wide-and-slow approaches that hold the potential for substantial aggregate bandwidth by forming many parallel connections running at reduced speeds, achieved via micro LEDs, VCSELs, or a combination of both.

Russell Ellwanger executive
#3

Right now we have several active engagements with Tier 1 customers, and those could bring significant volume in a new part of the optical communication market. As we said before, NPO is a market that's expanding for silicon germanium TIA and drivers, and it's generating customer demand at a level we've never seen—demand we're meeting today from three 200-millimeter fabs and one 300-millimeter fab. While we add that capacity, we're also continuing to advance our technology road map, and we're happy to report that design-in momentum is already strong for our H6 next-generation silicon germanium technology, which is in production for the most demanding products. Our technology's strength also got recent recognition when Semtech—our strategic silicon germanium partner for over 2 decades—earned an innovation award at ECOC '26 for its Optical Interconnect Solutions. In Agrate, Italy, our 300-millimeter capacity corridor recorded its highest RFSOI revenue quarter-to-date. Fab 10 likewise set new records for product tape-outs, which shows how strong our long-standing customer relationships are and how exceptionally the Fab 10 team executes in delivering the technology road map, extending design win momentum on our industry-leading 300-millimeter RFSOI platforms; we expect this trajectory to continue over the coming quarters. Tower's BCD technology demand continues to gain strength on both 200-millimeter and 300-millimeter platforms, showing roughly 10% growth year over year. The drivers behind this are not just the fast expansion of AI and cloud data centers, but also rising customer success in mobile and automotive applications. That momentum is visible in growing customer commitments, as several strategic accounts have signed long-term agreements at higher ASPs, which leads to forecasted blended prices for 2027 that are 20% higher than in 2026; meanwhile, in CIS and displays, production has started on a new high-speed global shutter sensor for high-end video cameras with one of the DSLR MILC industry leaders. Working alongside our key customers, we're fueling the next stage of business expansion, generating demand ahead of the considerable added capacity that the Uozu expansion will bring online; we're likewise collaborating intently with our OLED on silicon display clients to supply capacity for the sharp ramp they anticipate in 2027. Moving to utilization: throughout the third quarter, every one of our 200-millimeter fabs operated within an 85% to 90% utilization range. Our 300-millimeter 12-inch fab is still fully utilized, comfortably exceeding our 85% utilization model. And as we noted, considerable capacity is being introduced across all sites, consistent with extraordinarily rising demand. In short, we began the latter portion of 2026 carrying robust commercial momentum, underpinned by widespread client demand spanning our distinctive technology platforms and persistently elevated usage throughout all of our production capacity. Crucially, the vigor we observe reaches past this quarter—considerably further out. Interactions with customers, engineering work, and clarity on demand keep strengthening our assurance regarding forthcoming expansion prospects, made possible by the exceptional, substantial worth that every one of our specialized analog technologies delivers. The prospects before us look very powerful and are in fact getting better, given that demand has climbed in each of our main business divisions, with sales growth expected throughout 2027. SiPho and silicon germanium are growing at a rate never seen before, and we have captured most of the pluggable scale-out market; we benefit not only from this segment's considerable expansion but also, through tight partnerships and cooperation with the market leaders, from robust developments that enlarge our served market, as we aim to keep spreading our differentiated technologies into multifaceted applications. The count of advanced NPO engagements is merely one among many signals of the strong partnerships that are capturing the future. To sum up, for Q4 '26 the SiPho annualized run rate should surpass $1.1 billion, with meaningful further target increases in Q1 '27. We will finish this year in an extremely exciting position. At this point, I'd like to turn things over to Mr. Oren Shirazi, our CFO. Oren, the floor is yours?

Oren Shirazi executive
#4

Thank you, Russell. Hello, everyone. In the third quarter, Tower's financial results took yet another notable jump forward, delivering record revenue, record gross profit, and powerful operating leverage throughout the P&L. Revenue rose 36% year-over-year, reaching $526 million, and gross profit climbed 92%. Operating profit grew 2.5x, and net profit came in at $120 million — a 23% net margin, or 2.2x compared with the third quarter of 2025. What these numbers show is that our accelerating growth is converting into considerable expansion of earnings. They also speak to the growing share contributed by our differentiated, higher-value technologies and to how well our operating model performs as we scale. Importantly, we begin this next stage of growth from a place of financial strength, as shown by our solid balance sheet and cash position; long-term customer contracts that underpin our CapEx investments will deliver fast ROIs by giving us greater visibility into future customer demand and high capacity utilization. I'll now walk through the quarterly financial performance, then cover our balance sheet, cash flow, and capital investment plan, and finish with the financial framework that underlies our long-term business model. Turning to the P&L. Third-quarter 2026 revenue was $526 million, the highest in the company's history, up 14% quarter-over-quarter versus $460 million in the second quarter of 2026 and $130 million higher year-over-year, a 36% increase year-over-year. Gross profit in the third quarter of 2026 hit a record $179 million, up 92% from $93 million in the third quarter of 2025. Operating profit for the third quarter of 2026 was $127 million, which is 2.5x the operating profit recorded in the third quarter of 2025. Net profit for the third quarter of 2026 was $120 million, a 23% net margin. That net profit is more than double the $54 million posted in the third quarter of 2025. Earnings per share in the third quarter of 2026 were $1.06 basic and $1.05 diluted, versus $0.48 basic and $0.47 diluted in the third quarter of 2025. Now to our balance sheet. Prepayments received from SiPho customers, mostly toward 2027 capacity reservation — including the $290 million of prepayments we announced previously, received in the first quarter of 2026 — appear on the balance sheet as liabilities under short- and long-term customer advances and are included in the 2026 cash flow report as cash flow from operating activities. Our balance sheet remains very strong, as the following indicators and financial ratios show. At the end of September 2026, our assets totaled $4 billion, made up mainly of $1.8 billion in net fixed assets, predominantly Fab machinery, and $2 billion of current assets. Shareholders' equity reached a record $3.2 billion, which is 80% of total assets. Our current ratio stands at a strong approximately 3.7x, and the increase in our long-term asset balance mainly reflects the Newport Fab lease prepayment as well as other prepayments. On hedging, I'd now like to describe our currency hedging activities. On the Japanese yen: TPSCo books most of its revenue in yen, and nearly all of its costs are in yen as well, so for the bulk of our Japanese business and operations we enjoy a natural hedge. To soften part of the residual yen exposure, we enter into 0 cost cylinder transactions as a hedge against currency swings. So even if the yen-to-dollar rate moves, the effect on our margins stays limited. Turning to the Israeli shekel: we earn no revenue in that currency, yet some of our costs in Israel are denominated in shekel. We therefore hedge a substantial share of that currency risk as well, using 0 cost cylinder transactions to reduce the exposure. So although the shekel-to-dollar rate can move, as it did over the last few quarters—a period in which the shekel strengthened sharply against the dollar—the effect on our margins remains limited. Now to the CapEx investment plan. As we announced previously, to support rising SiPho and 5G demand we are carrying out a $920 million investment plan aimed at expanding the capacity and capability of SiGe and SiPho equipment across our 8-inch fabs in Israel, Newport Beach and Texas, together with our 12-inch Uozu fab in Japan. These investments are running ahead of schedule on on-schedule tools, technology, process qualification, equipment arrivals and ramp plans, and customer contracts and prepayment support them. Roughly 65% of this $920 million CapEx investment has already been paid and sits within our cash flow from investing activities for the reporting period through the third quarter of 2026, while the remaining 35% should be paid out in Q4 '26 and H1 '27. As Russell mentioned earlier, we expect a 1 year full ROI on this investment; beyond that, we are moving forward on the Track-1 Japan 300-millimeter CapEx plan by issuing purchase orders for 12-inch SiPho capacity and capability tools, which will substantially raise our 12-inch SiPho capacity. Those CapEx payments are slated for 2027 and 2028. On the business model: as announced recently, the company has updated its business model, which now incorporates the planned investment in the Arai facility, formerly Fab 6, to repurpose it for 12-inch SiPho wafer manufacturing, thereby maximizing the company's 12-inch wafer output in Japan, with support from METI, Japan's Ministry of Economic, Trade and Industry. The updated model rests on forward-looking operational business and financial assumptions, among them the assumption that every fab will run at 85% utilization after full installation and qualification of the $920 million in SiPho and 5G CapEx investments. We keep ramping SiPho and 5G capacity and capability in our Fabs 2, 3, 7 and 9; beyond that, we are working to repurpose the Arai facility and to qualify the new equipment tools to be installed there. We're making assumptions about forward-looking wafer selling prices for products that exist now and those coming in the future, along with flow, plus forward-looking assumptions on cost and completion dates for successfully qualifying the tools we plan to install, the process technologies to be qualified, and customer products to be qualified and ordered from the company. Under this updated financial model, we assume a target of $3.6 billion in annual revenue, which comes in $760 million above the February model target and reflects a 27% revenue increase. That gives us $1.63 billion in annual gross profit, which works out to a 45% gross margin versus 39% in the prior model. This annual gross profit sits $510 million above the previous February 2026 model, representing 67% incremental gross profit coming from the incremental $760 million in revenue and reflecting our enhanced product mix. We're also targeting $1.38 billion of annual operating profit, which yields a 38% operating margin compared to 32% in the prior model. And we target $1.2 billion in annual net profit, resulting in a 33% net margin versus 26% in the prior model, 20% in the second quarter of '26, and 23% in the third quarter of '26. This $1.2 billion annual net profit target is $450 million above the pre-February 2026 model, representing 59% incremental net profit derived from the incremental $760 million in revenue and reflecting our enhanced product mix. That wraps up my prepared remarks. I'd now like to hand the call back to the operator so we can take your questions. Operator?

Cody Acree analyst
#5

Congratulations on the progress — it's really impressive to see. Russell, maybe you could help me with the math behind those exponential capacity-addition figures you provided. Could you just level-set us now that the $920 million CapEx program is firmly on track and actually ahead of schedule? And can you level-set, from a revenue standpoint, your fiscal '28 capacity support for SiPho and SiGe?

Russell Ellwanger executive
#6

I'd rather not provide an exact revenue target, but I can give you a multiplication against last quarter's revenue in terms of raw capacity. As I said, Q3 delivered $240 million in SiPho revenue; we'll end Q4 at 2x the starting capacity that produced those Q3 revenues. So, 2x the starts behind the Q3 shipments. That revenue turns into the run rate in Q2, once a full Q1 at that start rate has shipped — December, January, and February all shipping inside Q2. Then in Q3 we add another 1x, meaning 3x the Q3 shipments, through the expansion under way at the Fab 7 Arai factory; that comes to roughly the SiPho capabilities from investments we've already announced. We're evaluating, and will very likely announce shortly, further SiPho announcements — we've identified additional expansion areas, mainly at 2 of our 200-millimeter sites. We haven't released that yet; we haven't discussed either the CapEx amount or how much capacity it would add, but it's another meaningful tranche on top. Still, on capacity, applying the multiplications I mentioned, you'd get something like a $2.85 billion to $2.9 billion SiPho run rate as of Q3 in start rates; that said, it isn't really 100% accurate, since it excludes what we're doing with integrated laser. That sits outside and on top of this, and is a substantial forecast in its own right. So if that gives you a baseline for SiPho revenue, for everything announced and fully implemented on starts — by plan as of the third quarter of 2027 — it lands somewhat above the $2.8 billion run rate.

Cody Acree analyst
#7

Thanks, Russell — that's genuinely useful, and a solid starting point for Q3. Given everything you've laid out, it seems like the summer update to your business model — the $3.6 billion and $1.2 billion in net profit Oren just walked us through — might already be too conservative. Would you say so?

Russell Ellwanger executive
#8

I won't comment on that. No — I'd call it an aggressive target. Still, we would never hand the Street a target unless we had very, very strong confidence that we'll hit it.

Cody Acree analyst
#9

Lastly, if I may, Oren: this quarter you reported a very, very strong incremental margin — the first time we've seen it in the 60s, at 62%. How sustainable do you believe that level is? Is it a new baseline, or will it move around as new tools come in and you begin depreciating that capacity?

Oren Shirazi executive
#10

Yes, precisely as you described. This quarter, not all of the tools had yet started their depreciation schedule — from what I said about 65% of the tools being paid, so you can perhaps gather that roughly 65% of those $920 million in tools began depreciating and the rest did not, and then there is the second tranche of Track-1. As a result, more depreciation will hit us down the road. That's the reason our long-range framework, in my view, doesn't carry the 65% incremental gross margin posted this quarter, and instead typically sits around 55% to 58%. Accordingly, we're reiterating our guidance of 55% to 58% incremental gross profit.

Vijay Rakesh analyst
#11

Russell and Oren, I have a question about the SiPho side. Clearly, you're offering several variants of NPO — external laser and internal integrated lasers. Could you discuss how pricing and contracts are shaping up as you consider some of the newer configurations and NPO shipping out? And one quick follow-up.

Russell Ellwanger executive
#12

That's a really, really intriguing question, and a good one. Pricing targets are part of our long-term contracts. Occasionally there have been some inputs that limit price increases, but the effect is the reverse. When new entrants possibly arrive in the game, one route they attempt to enter the market is through deep discounting. Those who discount deeply usually do so because they can't compete on technology; but what our contracts accomplish is to establish a starting bar, not that if the market truly called for some price reduction we wouldn't agree to be reasonable with our customers; but our pricing is set very, very effectively in our long-term contracts, and pricing targets exist within all of our contracts. For NPO, I'd rather not provide an exact figure. I don't believe that would be the appropriate thing to do. But NPO definitely follows our pattern that any new platform resets a pricing curve in terms of achieving a higher ASP. So the manner in which anyone safeguards high margins over time is by being quickest to deliver products and capitalizing on being first in the market to help a customer expand their market by being first with their particular end application; so I'd simply like to say that our pricing is solid, it's good, and it's very sustainable given how we have structured our long-term agreements. NPO will start from a higher point than a pluggable, and that makes sense since it's a more advanced technology platform. Our pluggable starting point wasn't bad and isn't bad at present, as the incremental margin drop-down shows. If that...

Vijay Rakesh analyst
#13

And then on the — yeah. That's really useful. I think one thing that kept coming up at OCP US was, clearly, folks are beginning to discuss 3.2T, 400 gig per lane more. When you look at your NPO demand backlog, are you seeing more engagements beginning to ramp up at 3.2 gig? Does that mean a bigger dollar figure and a more complex layout on SiPho? How do you view that mix and the opportunity?

Russell Ellwanger executive
#14

Yes is the answer to both parts of your question.

Vijay Rakesh analyst
#15

Is the mix higher on 3.2T?

Russell Ellwanger executive
#16

Yes.

Simon Coles analyst
#17

Given how strongly you folks are performing, there's plenty of talk going around regarding competition on SiPho. I believe you began ramping 12-inch yields late last year, early this year. Could you give us an update on that progress and on how it's helping you hold onto your leadership position?

Russell Ellwanger executive
#18

I'm not certain I fully grasp your question. Let me respond to what I believe you're asking. If you need more clarification, just let me know. We did begin shipping revenue volumes in the first quarter of 2027, at extremely high yields — really high yields; with our 300-millimeter platforms, we have very special applications; we have an amazing demux capability that I believe is absolutely the best anywhere in the world. So we've added several new types of platforms. Most of the NPO tape-outs we're doing are being done at 300-millimeter. And those are 16 lane, 32-lane applications right now. And the pluggable we're doing also delivers just incredible performance, substantially better on insertion loss than I think anyone else can even come close to. So if that answers your question, the ramp we're doing at 300-millimeter is quite substantial. We have stated that from where we'll be in Q1 on 300-millimeter, which on an equivalent basis is somewhere around 10,000, 200-millimeter. By the end of the year, we'll be increasing that — by the third quarter of the year, we'll be increasing it by 3x and then incrementally increasing as well in 2028 and ultimately, a very, very big increase as the new shell is built at the end of 2028. Did that answer your question? I hope it did.

Sahej Singh analyst
#19

Russell, the way you laid that out really helped, so let me pick up from there. You mentioned that in 3Q, SiPho revenue came in at roughly $240 million. That contributes — and the step we take in 3Q '27 puts another 1x of capacity on top of it, thanks to Fab 7 and Arai.

Russell Ellwanger executive
#20

That makes it three times as much.

Sahej Singh analyst
#21

Sorry.

Russell Ellwanger executive
#22

Triples it.

Sahej Singh analyst
#23

Sure, triple it — though by my calculation that comes out to about $960 million annually. Which means it's $200 million above the $760 million that Track 1 contributes to the '28 model. So first, if I've got that right, is the whole thing Track-1? Or are we also counting leftover capacity from the $920 million program? And second, using those same units, as Track-2 ramps up toward quadrupling Japan's 300-millimeter, roughly what multiple does Track-2 bring? Is it another 1, 2x? Or is the floor higher than that?

Oren Shirazi executive
#24

On your first question, what Russell said is that as of today — I mean, right now, total revenue from Q3 shipments comes to $240 million, and tripling that gives $720 million, not $960 million. So that's a $720 million per quarter run rate heading toward — I believe this was Cody's answer — the question about 2028. So Russell said the Q3 '27 capacity baseline can support capacity of $720 a quarter. Multiply by 4 and you get $2.8 billion, which is what Russell mentioned, right — not $960 million.

Sahej Singh analyst
#25

All right, that's useful. And then regarding...

Russell Ellwanger executive
#26

I did mention as well that there are extra SiPho activities not counted in that figure; so I'm not claiming we're capped at $2.8 billion. The integrated laser is in there, lots of things aren't part of that number. If we're speaking specifically about the quadrupling, what we're quadrupling is the current capacity we're discussing, which sits at roughly 10,000 wafers per month in Fab 7 300-millimeter, and that gets quadrupled to approximately 40,000 wafers per month. I mean, we don't yet know the precise figures, the build hasn't been completed. But what that gives you is a huge increase if it were all SiPho, since you're putting 2.25 for every 300-millimeter wafer when compared with 200-millimeter. So if you went to 40,000 in 200-millimeter equivalents, you're near 100,000 wafers.

Sahej Singh analyst
#27

That helps a lot. So, Russell, on that P times Q math you're referring to the 2.25x increase in die size on the 300-millimeter — that makes sense; but Track-2 is also going to be mostly NPO, which means the content uplift is there too, and I think that shifts the like-for-like math that most investors are trying to do. Could you give us some color on that as well?

Russell Ellwanger executive
#28

The price per wafer, again, had been stated as we move to NPO, it resets the price target, right, it resets the pricing curve. What bigger chips do is they create more silicon demand. So the NPO will — the more lanes that you put into it, the more silicon you're going to need to ship the same amount of chips. That physically changes the cost of the silicon other than the higher the integration, the newer the technology, the higher the starting price point.

Sahej Singh analyst
#29

Right, that's it. I mean, we're aligned on that. I think the main thing is really about sizing it bigger. But let me squeeze in one final question, and it's probably about timing. I believe you mentioned Track-2 is well underway, and I thought Fab 7...

Russell Ellwanger executive
#30

No. What I said was that Track-2 will be well underway in 2028.

Sahej Singh analyst
#31

'29 — '28.

Russell Ellwanger executive
#32

Starting in '28 with the buildings, we'll move on in '29 once our final contracts, which we're in the process of signing right now, are done. Once again, we pushed our capacity agreement out to 2029, and that is a really, really positive development for us — huge visibility and dedication; so we had mentioned the 1-year ROI on that $920 million investment made mainly for the 200-millimeter expansion. Having those customer commitments genuinely strengthens your confidence in the short-term ROI. And the ROI we're — the model is extraordinarily aggressive.

Sahej Singh analyst
#33

On Track-2 specifically, I believe you're taking ownership of Fab 7 on April 1. Is there a chance that could begin sooner—meaning, to put it simply, might Track-2 output arrive in the second half of '28? Or does it essentially all depend on '29?

Russell Ellwanger executive
#34

It could be possible. What we truly believe, though, is the build time involved, and we're fairly confident that '29 will be the expansion. If it can be sped up by a quarter, that is definitely what we're aiming for, correct? I mean, with any investment, the sooner you can start shipping, the sooner you satisfy your customers and the sooner you satisfy your investors.

Analyst analyst
#35

Sebastian here, covering for Mehdi. I suppose I'll build on what was asked about the competitive landscape looking ahead to 2027. With one of your big rivals ramping up PIC capacity, how ought we to view market share in silicon photonics and silicon germanium for 2027? And I have a follow-up.

Russell Ellwanger executive
#36

I'll turn it over to Dr. Racanelli to address that, since he works directly with driving the BUs and has a tremendous amount of interaction with customers. Then, if there's anything I can add, I'll gladly do so.

Marco Racanelli executive
#37

Thank you, Russell. It's always difficult to pin down market share precisely, especially in a market expanding as fast and as swiftly as this one is, on both the silicon photonic and silicon germanium sides. And when a market is expanding at such a rapid rate, there's clearly space for more than a single participant, and that's exactly what we're observing, as you mentioned, a bit of competition. What I can tell you is that we have — we were the first to ramp in 200-millimeter to very high, not only capacity, but also very high yields with very good performance. And as we discussed earlier, we were also the first to ramp in 300-millimeter with very high yields and very high quality and performance. So that gives us confidence and gives our customers the confidence to extend the LTAs, as we discussed, into 2029. So we're building capacity as fast as we can, and all of that capacity gets claimed almost right away once we announce it, certainly well before it's even in place, through contracts from customers. So that's a tremendous position to be in; we also keep investing at the leading edge of technology alongside our customers, having many Tier 1 customers with us; we have incredible visibility into their future needs; and so we're putting in place technology that will pay dividends well into the future and not just today. So we expect not only to maintain very good market share in the short term, but certainly, there will be others competing with us in the market; and with growth as it is, that's perfectly understandable and fine; but also in addition to the short term, we're putting technology in place — 3.2T was discussed with both silicon indium phosphide, lithium niobate technologies and CPO, as we've mentioned, both on the silicon photonics side and also silicon germanium side — putting technologies in place that will enable, at least at the leading edge, us to maintain our competitive advantage and market share for the long term.

Russell Ellwanger executive
#38

Thanks, Marco. I’ll just add one more thing, and I believe this is a really important point to include. We are genuinely involved in long-term agreements worth billions of dollars. Customers are never going to commit themselves to the end of ’26 at this stage — not ’27, not merely ’28, but 2029 — unless they are confident that, first, you can deliver wafers on schedule and at the highest quality. And second, that you will keep a differentiated road map that is more than competitive. Most of what we ship today to lead customers will represent only a fairly small share of what we will be shipping in 2029. So the confidence customers have from our relationship and from the programs we are working on — programs that are not only current and next, but generation plus 2 and, in some cases, generation plus 3 — that is what gives us assurance. Competition will exist. It’s a large market. It’s a good market. It’s large enough to support competition. Still, we are very confident about preserving our leadership position, because we are leaders working with leaders on leadership programs. And I think that is the most important thing I can say. As for what our market share will be — at present, I think it is extremely high, and one could verify that with revenue. Just look at Q3: close to a $1 billion run rate on SiPho revenue. I think you would be hard-pressed to find others who can come close to that. But with Q4 breaking $1.1 billion, we are in a pretty substantial position. Yet the most important thing is not the market share we have now; it is customers’ confidence to sign long-term agreements through 2029, because those are committed; those are committed capacities; and they have to know us. We have to know them. A partnership and trust must exist in order to enter into those contracts. And that, for us, is what really gives us confidence in maintaining a very high market share: we are working truly with the best of the best on very advanced programs that — this is a joke I have, but it is really true. I mean, for Kevin Costner in the Field of Dreams, “if you build it, they will come” may be true. For us, we need very strong customer partners. And if we have the best of the best in each market, and we are working with them together on next-generation and generation plus 2 technologies, when we build it, it will be sold, because we are dealing with the leaders. So again, perhaps enough has been said; but to give a specific market share number — I know what I think it is right now — to say what it will be at the end of ’27, ’28, ’29, that would really be proctology. That said, claiming we'll hold onto a strongly differentiated market share in products with significant technological differentiation—I think stating that takes a lot of confidence.

Analyst analyst
#39

And my follow-up, I believe, was about a different end market, RF mobile. My sense is that RF mobile is getting close to the bottom. This quarter it sits in the low 50s. With the tape-out activities you brought up on the prior call, how should we think about where RF mobile goes as we head into 2027?

Russell Ellwanger executive
#40

You mentioned what, sorry — could you repeat that?

Analyst analyst
#41

Sorry, when it comes to activities, how should we think about the trajectory for RF mobile?

Russell Ellwanger executive
#42

From our perspective, that trajectory looks very strong. We experienced a bit of a dip as we moved from Fab 7 to Fab 10, since we needed to requal multiple customers because Fab 7 was really repurposed to focus on optical, between the SiPho and the SiGe; it's going very well. We expect Q2 — while Q1 was the highest shipments we've had, Q1 will be — Q3 was Q4, Q1, Q2 should keep growing, with Q2 being absolute records; so I think it's — for us, it's going extremely well. Again, I don't have a specific feel for the exact market share percentage; we are working with leaders and growing market share with the leaders; and so I believe we're in very good shape. I'm also quite confident that one of our programs isn't just a step function in capability; it's an absolute opening of a new market capability itself. And that has not been formally released. We're working with at least one very lead customer on that type of figure of merit, and it's truly a breakthrough figure of merit. Marco, is there anything else you'd like to add there?

Marco Racanelli executive
#43

No, I think that came from you. In SiPho and RFSOI, we lack the share we hold elsewhere. Several competitors exist, yet at the very leading edge we remain on top; so our share is typically concentrated in that very leading edge, the cream of the crop, so to speak, and that is where our focus lies. Capacity for RFSOI is limited on our side. At this point Fab 10 has been dedicated, and we employ it for those leading components and leading technologies.

Richard Shannon analyst
#44

Nice work once more on the strong numbers. Russell, I'd like to pick up on a few things you mentioned during the call about SiPho — particularly the laser attached side, which isn't included in the revenue model you've discussed. Could you perhaps address it from a few angles, one being how much content add there is with laser integration or laser attach here? And as time goes on, what share of your PICs do you figure would carry laser attach, and, well — I believe there are several approaches one could take. What attach rate do you anticipate over time?

Russell Ellwanger executive
#45

The market really moves in several directions. With integrated laser, we're entering a new served market that could be very large. Some analysts claim integrated laser will be a massive portion. Others simply don't cover it. I honestly don't have a definitive answer. Our lead customers in that area are extremely committed to growth, with very strong forecasts. And once more, those figures aren't in the model yet, but they could be quite appreciable — particularly since, as one might imagine, ASPs are appreciably higher. Against what is already a very good baseline. So it's an attractive market to enter. Die attach, however, serves more than just the indium phosphide laser. It also covers an indium phosphide 400G modulator that we've more or less already demonstrated, plus lithium niobate modulators. So the die attach work we do spans many different things, and it could represent a very large portion of a 6.4T form factor for NPO or a pluggable featuring a 400 gigabit modulator.

Richard Shannon analyst
#46

All right, that's a useful way to look at it. Now I'd like to ask about NPO. You mentioned on this call that there are more than 10 customers. Could you perhaps describe the kinds of customers involved here — across pluggable transceiver suppliers, the historic players, OEMs, and hyperscalers too? And then, how would you describe the competitive dynamics in this area? Finally, it would be great to hear any qualitative thoughts on the market share you anticipate in that product category over the next couple of years or so.

Russell Ellwanger executive
#47

For pluggables, we already have a set of customers, and every one of our NPO customers comes from within that group. So there is nobody we're doing NPO with whom we haven't already done pluggables. It's an extension, or a newer capability, aimed at a broader market. So it's not a different customer set at all. One reason I think we have so much engagement is, as I said, the confidence and trust that has been built up through our work with pluggables — so that covers the first part of your question. We expect an appreciable amount of NPO to ship in the second half of this coming year. That's what we expect and what we're targeting. Will it actually happen? That really hinges on market adoption. But it's the same group of customers simply running more advanced programs with us, if you will — as I mentioned earlier, Gen plus 1, Gen plus 2, and Gen plus 3 programs are readily available between us and our premier customers. Marco, is there any color you'd like to add there?

Marco Racanelli executive
#48

Yeah. No, I think, as you mentioned, how large NPO becomes for us over the near term will hinge somewhat on how the market adopts it. What's great for us is that we hold a strong position in both pluggable and NPO. So if it isn't NPO and it's pluggable, that's still Tower wafer being built. So — and yes, I think, as Russell said, the majority of our current customers are the ones developing NPO products with us. We haven't made that public specifically with them, but we'll likely aim to put out some PRs so it's in the public domain before long.

Russell Ellwanger executive
#49

Thanks very much. I appreciate the questions. To wrap up with a closing remark: we established a corporate theme — and it remains our corporate theme. It took effect in 2025 and consisted of bold growth, limitless impact, infinite reach. We truly believe in that and strive to embody it in all our work. Being bold in growth, believing limitless impact is achievable, and pursuing infinite reach genuinely inspire the customers we serve as partners. We hosted a highly successful customer conference in Shenzhen, China. In roughly 1.5 weeks, we will host our customer conference in Santa Clara, California — the Technology Conference. It honestly felt like getting together with old friends — I really can't emphasize that enough. When you stick to — and we genuinely make an effort to stick to — our values, we have these things we refer to as our value vectors: leadership, partnership, impact and innovation. When you stick to those values and what they mean — I mean, the real fundamental meaning of leadership, I think, is captured in something John F. Kennedy said, that leadership and learning are indispensable to each other. If you stick to partnership, which means genuinely working side by side with your customers to adapt yourselves to what the market demands, and to do it quickly, really quickly, it produces amazing relationships as you achieve success together. That's what we see through our long-term agreements — a customer acknowledgment of the success we've shared. We get a lot of joy from it. They get joy from it, since success begets success, and that can never be overstated. Impact matters enormously; and impact really means giving your people the power, the capability and the ownership to make change, to make things happen. Many of our customers are amazed at how many young engineers they interface with, because we focus on developing very high-talent, high-potential people. And their growth really depends on working with customers who inspire them. There are 3 things that truly enable success: capability, environment and, genuinely, motivation — your passion. Provided you bring enthusiasm and the right skills, and you build a collaborative atmosphere in which the client feels comfortable speaking candidly—sharing criticism about where you need to improve without you becoming defensive, while you can likewise point out what they should do better—think about what that relationship turns into, and how genuinely joyful and motivating that period is. This is also where we discuss impact and innovation, which matters enormously, and our approach has been, and remains, that the gap between gross margin and operating margin is very, very small. As I've mentioned previously, in the $3.6 billion model we released, gross margin was 45% and operating margin was 38% — a 7-point decline from gross to operating margin. Why is that so significant for us? We aim to keep infrastructure lean because that enables speed. And the faster we go, the more we delight our customers. The questions — not whether, but the questions about NPO, the questions about market share later on — those are extremely, extremely critical. But how do we ensure it happens? Speed of implementation matters here, and that aligns with how we approach innovation—since occasionally the most significant inventive step is making things simpler, which reduces the gap between gross margin and operating margin. That said, I'd like to wrap up the call. Our current position and future direction are genuinely thrilling. I'm grateful to our investors for their engagement with the company. I deeply value our customers' confidence and trust, as shown through the LTAs. And our team members across the globe are outstanding. They contribute continuously and receive compensation that I believe matches their efforts, yet what matters most to them is the satisfaction of heading home each evening knowing they've achieved something and having had the resources needed to succeed. We're eager to meet with you at the conferences coming up on our schedule. First, the Barclays 24th Annual Global Conference takes place in San Francisco on December 10; then Nomura at CES 27 in Las Vegas on the 5th and 6th of January 27, followed by the 29th Annual Needham Growth Conference in New York on the 12th and 13th of January; and, as always, analysts and investors are welcome to reach out to our Investor Relations department whenever questions arise or you'd like to follow up. Your feedback is something we truly value, since it helps us improve. Many thanks.

Operator operator
#50

That brings us to the end of today's conference call. Thanks for joining us. Everyone can go ahead and disconnect now. Have a great day.

Read the full transcript via the API

You're viewing the first half of this call. Get the complete Tower Semiconductor Ltd. transcript - plus 131,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.

Get an API key View API docs →

For developers and AI pipelines

Programmatic access to Tower Semiconductor Ltd. earnings transcripts and 131,000+ others is available through the EarningsAPI REST API and the hosted MCP server. Quarterly plans from $162 - full transcripts, speaker segments, full-text search, and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.