Okay, we'll go ahead and get started. Hello, everyone. Good afternoon. My name is Quinn Bolton. I am the Semiconductor Analyst for Needham & Company. Thank you for joining us at the 26th Annual Needham Growth Conference. It's my pleasure to host this fireside chat for Navitas Semiconductor. Navitas was founded in 2014, and is the only pure-play, next-generation power semiconductor company in the semi industry. The company's GaNFast power ICs integrate power, drive, control, sensing, and protection in a single monolithic die, while its GeneSiC power devices are optimized for high power, high voltage, and high reliability. Joining me from the company are Ron Shelton, Chief Financial Officer, and Stephen Oliver, VP of Corporate Marketing and Investor Relations. Ron, Steve, thank you for joining us.
Yep, you bet.
So I'm gonna start, Ron, put you right on the spot. Yeah, we saw you last week at CES.
Yeah.
The next day, the company made an announcement about a CFO change.
Uh-huh.
What, what, what can you tell us about the CFO change?
Yeah, sure. It was the next day, right?
It was the next day.
When I talked to you. Yeah, yeah, yeah, yeah. No. So, so, when I.
Makes it a head job.
When I joined the company about two years ago, so when I joined, the arrangement was always, and the agreement was basically a two-year tenure. And I had committed to two years, and in that two-year period, the idea was, you know, the company was newly public at the time, help it mature as a public company, build out a finance staff. You know, during that time, we did three acquisitions. We raised $100 million. I think from when I started to today, you know, we've developed great relationships with analysts, investors like yourselves, and the Wall Street community in general. So that was sort of my objectives when I started two years ago, and that's kind of where we are now.
And it was the perfect time for the transition. You know, I'm signing the 10-K, so just to put everyone at ease, that's not why I'm leaving. It's a well-planned transition. My successor, Janet Chou, is from WD, Western Digital, most recently. She's got a ton of experience in semiconductors and related fields. She was with NXP in China. She was CFO of a company, JCET, which is an OSAT in China, that was publicly listed in China. Most recently, she was CFO of Western Digital's operations group. And as you know, Western Digital is multi-billion. So, I've not spent a ton of time with her yet. I will in February and introduce her to the community, but she's fabulous.
So she's really, really good, and I like her a lot, and I think from a transition standpoint, it's probably the easiest transition I've been through in my career. She's really good. So super excited for the company to have her join, and she's there now, and she'll assume the CFO role full-time on March 1.
Great. Well, we will certainly miss working with you and look forward to meeting Janet, you know, over the next coming months. The other thing that was announced last week at CES, well, maybe not at CES, but CES week, was another acquisition in the GaN space. Renesas coming in and taking out Transphorm. Do you guys have any thoughts on that transaction? Do you think it changes the competitive landscape in, you know, the gallium nitride market?
Perhaps.
I think it's a, sort of a great sign that it's a market that people want to get into, so that's a good thing. I think, Transphorm, a great team, entered the GaN market with an early form of GaN, that didn't quite get traction. The Navitas version obviously is the later form that's easier to control and can be integrated. So there's much more value there, especially now that we have, the GaNSafe portfolio that can break through the glass ceiling for this new technology into higher power, more reliability-conscious settings. I think, maybe over to Ron to comment on the price paid for the company.
It was a good price for them. I'm happy for them. Yeah.
Nicely over 10x revenue, right?
Yeah, yeah. No, it was, it was a good price. I think, I think it's. We don't, we don't really see them. I think we compete in different parts of the market. You know, so, so again, I think, I think they got a good price. They should be happy.
Great, we'll turn to you know, kind of more Navitas-specific questions. Let's start with the demand environment, inventory situation. You know, while the broader semi market has clearly suffered from inventory digestion this year, Navitas' results were pretty strong.
Yeah.
You guys clearly outperformed the broader markets. What do you really attribute this, this outperformance to?
Yeah, I think I. You know, it's just the thesis that we've been discussing, you know, talking about over the last few years in terms of GaN specifically, and our advantages in the GaN market, I think became pretty apparent over the last couple of quarters. So, where others in our space saw weakness, and even what they perceived as being demand was really replenishing the channel. What we saw, again, is in the higher power charger market, specifically, where we have the majority share, we saw very strong demand there. We think the market is headed that way. That's where we're well-positioned, that's where our advantages show up. So our channel inventories are actually very low right now, especially for those products.
Yeah, just took my next question. So it sounds like despite the pretty strong revenue growth over the last few quarters, you feel pretty good about where customer and channel inventories are, especially in that.
100%. Yeah. Yeah, yeah.
You know, can you discuss sort of just order backlog trends? You know, they've been strong sort of Q2, Q3. Did that momentum carry through the fourth quarter, or would you rather, you know, push that to the fourth quarter earnings call?
Well, I think in the call coming off Q3, we talked about Q4 and going into the quarter being you know nearly fully booked. So yeah, I mean, so we were out there publicly with that. So going into Q4, it was very strong.
Okay.
Yeah.
Good. And lastly, just kind of on the, you know, kind of business environment, lead times, how have they trended or how did they trend in 2023? Are you kind of back to pre-COVID levels? How are where lead times looking?
Yeah, go ahead, Steve. I'll let you.
Yeah. So, in general, lead times, obviously, they vary by technology. So, for gallium nitride power IC, it's about 16 weeks with a, with a completely empty line. We've got great capacity commitment from TSMC, our foundry partner, of, 3x what was the 2022 baseline. For the silicon carbide side, through X-FAB in Texas, we've got a 5x capacity guarantee. For silicon carbide, one of the diodes is about 16 weeks. For a MOSFET, it's about 26 weeks. So that really hasn't changed. That's a, just a function of processing time.
Okay, perfect. You guys had a lot of new announcements in the second half of the year, announcing four new major technology platforms. You had GaNSafe, the GaNSense half-bridge , the third generation of the silicon carbide FETs, and then your first or the industry's first bi-directional GaN technology. What's important about those? You know, maybe starting with the GaNSafe.
Yeah, GaNSafe, basically, gallium nitride is a new technology. The first place you adopt any new technology is a place with relatively low risk. With a mobile phone charger, if you break it or you lose it, you borrow another one, or you buy another one. But for a high reliability, high uptime system, like a data center or for a solar microinverter, where it needs to be on your roof for 25 years, there's a certain credibility, there's a certain reliability requirement. So now that we've got 125 million units shipped, check. Now that we've got great quality in the field, check. We've got a 20-year warranty on our products, check. And with the GaNSafe, we've now developed a product pretty much on steroids. It's higher power capability. It has improved protection features.
This is a true IC, so we can do current sensing without energy loss. We can then detect and protect against a fault 6 x faster than anything else because we have an autonomous, logic-driven device. So we've got this incredibly robust thing, which also happens to be higher thermal performance, better mechanical integrity. So now we've got something that basically breaks through the glass ceiling for this new technology into data center, into solar, and into EV.
The GaNSafe is really kind of the key to breaking into those high-power markets that, w e'll get into some of the TAMs and.
Correct.
P ipeline, here a little bit later in the talk. The other technology, and I'll, I'll limit it to two technology questions, but, ever since acquiring GeneSiC, you guys have sort of highlighted the company's, trench, trench-assisted planar technology. Tell us a little bit about that, what's unique, and, and what advantages does that bring, especially since, the company, I think, is really the only fabless, or yeah, near fabless, since you're, you're bringing epi out, player in the silicon carbide market.
Right. Great questions. Yes, so for a silicon carbide MOSFET, the traditional way to do it is a planar or flat structure, which is quite easy to manufacture, is nice high reliability, but it uses a large area of die to make a certain performance, so it's not cheap. The alternative is trench, where you literally dig a trench, for part of the circuit. It basically folds it down, so you save on die area, but it's more difficult to do, and it's not as reliable. So you, you've got two compromised technologies. In the middle, basically, it's trench-assisted planar gate, where we take the best of both worlds without compromise. So the GeneSiC technology means you've got a very small die, which is easy to manufacture and very high reliability.
When it comes to actual performance in the field, if you do a like-for-like performance, as we did in our Investor Day a year ago now, you actually find that the GeneSiC device runs 25 degrees Celsius cooler than a typical one from the rest of the guys in the market. In solid-state reliability terms, that 25 degrees reflects as 3x longer expected lifetime. It's a really good benefit.
Great. Wanted to move to some of the points, you know, coming out of the Investor Day in December. I think a key theme was the diversification away from mobile adapters or the consumer segment into some of these higher power markets that we talked about with GaNSafe. But the data center, industrial motor drivers, solar, EV mobility. Maybe starting with EV mobility, you kind of highlighted your pipeline is now about $400 million, up 34% in the last seven months. You know, what are some of the biggest opportunities there, you know, in the cars, roadside chargers, onboard chargers? You know, what's sort of driving that pipeline?
Yeah, that's a mix, again, by technology. So in the EV space in general, starting with the roadside charger, so the 300 kW super system, we've actually now got a new EV long-haul trucking spec as well. That's a 1,250 V rail or cable, that's up to 3.75 MW, which is a stupidly big number. And for that, it.
Right?
If you want a truck to just charge for half an hour instead of 12 hours to go a long distance, that's what you need. For that, you're really looking at silicon carbide. Got great high voltage, high power capability, and GeneSiC technology gives us the lead in high voltage. We're the only company that goes from 650 V silicon carbide to 6,500 V, so it's an amazingly flexible technology. We're shipping today into that market for the SK Signet is a Korean company that's got a great market position there.
Actually, SK Signet and ourselves, taking advantage of the Inflation Reduction Act, with the installation of roadside chargers across the country, and the U.S. administration has said, "If we're paying for it, 55% has to be built in the U.S." We're in a great position because we build silicon carbide in Lubbock, Texas. That's a great advantage.
Pretty soon, the epi is gonna be in.
That's true.
California.
That's true. The epi will have got three Aixtron machines going into Torrance, California, so that's a great expansion for us. In the vehicle itself, if it's an 800 V rail, that's a great silicon carbide play. If it's a 400 V rail, that's gallium nitride, potentially silicon carbide. And we can use these different technologies to mix and match. The fact that we have both means that we're agnostic as to which one is best, and we can leave it to the customer, or our own design centers, to decide what is best in terms of performance and cost, and size. So there are examples where we can actually have an 800 V charger that comes from AC. So an 800 V battery needs 1,200 V silicon carbide.
If you plug it into the AC, GaN's the best solution. So we have an 11 kW reference design, which actually has both technologies in it, and that's the best solution.
The market today for auto or EV and mobility, I think, is all GeneSiC or silicon carbide-based.
Today, everything.
Revenue stream.
Yeah.
When do you see the GaN, you know, you talked about 400 V battery, potentially, I think maybe the lower end of the onboard charger market. Yeah, where do you see GaN and when could GaN start to intersect easily?
That's next year, 2025. Yes, as we've had on schedule for a long time. As a reference point, Geely, the Chinese EV company, we originally started a joint design center with them in Shanghai, just with the GaN product. And then working with them, that's on track for 2025. At the same time working with them, they had a problem with an alternative silicon carbide provider. We very easily dropped in the GeneSiC part, ran really well, and that's a good win.
Great. I think maybe for me, the thing that surprised me most, coming out of the Analyst Day was the industrial and the appliance market, where I think a lot of these wins are, you know, motor driver based. But your pipeline there had increased 250% in seven months.
Yep.
Stands at $360 million. It's your second biggest opportunity. I guess I sort of thought nothing moved quickly in motor drivers, so what, what are you doing to, to sort of, you know, shake the trees in motor drives?
Right. Yeah, motor drives is a, i t's a boring market, but it gets very exciting when you see the numbers. Yeah, so when you have a motor, you basically have a series of six switches that in turn, fire and turn the motor to make it move. Traditionally done with silicon IGBTs at a very low frequency, like 8,000 Hz cycles per second. That means that things are very big, they're very inefficient, they have big heat sinks to go with them. What we've been able to do with our technology, with GaN, and especially the integration, is create tiny little half-bridge modules. So you have three of these little modules. It shrinks the whole system. We can run at high speed. You can now integrate the electronics into the same casing as the motor, and at the same time, throw away the heat sink.
One of the biggest failure modes in a motor is when it rattles and the heat sink falls off. By eliminating a failure mode, we can not only improve the efficiency and the cost of the motor, but keep it running as well.
Got it. At the Analyst Day, you highlighted, I think, some opportunities with large consumer appliance manufacturers, and to be honest, I don't remember all of them, but some of them sounded, you know, pretty substantial. You know, $10+ million to $20+ million opportunities. Maybe highlight a couple of those, you know, for folks in the audience.
Sure, yeah. So basically anywhere there's a motor, there's a need for power electronics. And even in a home, you've got a washing machine that's got a spindle motor to actually turn the drum. You've got a motor that then does the pump in and out. You might have a heater system, you might have an auxiliary power supply. So, you know, you've got, I'll just say companies like Whirlpool, like Miele, like Bosch, there are so many different applications within an appliance. And the European Union is doing a great job of raising efficiency standards again. So the old way of doing it just doesn't cut it. Even if it was completely free, it doesn't meet the specs, so you have to have something higher efficiency.
The lighter weight, the faster speed are all cherries on the cake on top of that.
Got it. Okay, great. How about let's move to solar and energy storage. That's another $250 million pipeline.
Mm-hmm.
Talk about your position both with GeneSiC and the string inverter market, and then with, you know, your GaN solutions and Enphase and the microinverter market. And maybe just, I think solar is a market that you've said is a little bit slower in terms of near-term demand. You know, maybe just address really what you're seeing broadly in the solar market as well.
Right. Yeah, so solar is one of the markets which is soft right now. There does seem to be an inventory of the old silicon-based product that we need to eat through. One example on the gallium nitride side is Enphase with the microinverter. This is a per-panel power converter. It's now up to 550 W electricity generated per panel. They've got a little bit of inventory with silicon, so we'll be shipping in the back half of this year with their new design, and they've been very public about going to GaN with their microinverters. So that's really a move to the future. Previously, they'd given us testimonials saying that, you know, they tried going from silicon to silicon carbide, just didn't make it work.
But by going to Navitas GaN, they could save a significant portion of the cost of their microinverter to do that, all based on the speed of the technology. We estimate that to be about 25% cost reduction per unit. For silicon carbide, that's shipping today with the GeneSiC product line. Companies like KATEK in Germany, Sungrow, Chint, a lot of different applications from 3 kW to, you know, 11 kW of power conversion. That market, again, a little bit soft as a macro. In general terms, our silicon carbide, we might have 2%, maybe 3% market share in the world. So even if the market dips, we can still grow based on that, and it's all about growth.
Just sort of increasing penetration rates and.
Correct.
Expanding your account base.
Yes.
Yeah. Okay. Data center, obviously, lots of questions, I think, you know, it seems like for every company, somebody's going to ask, you know, how do you benefit from AI?
Mm-hmm.
I'll ask the same question for you. The data center, is it just, you know, more power supplies, higher wattage power supplies? You know, what are you seeing in the data center?
Yeah, AI is for us a great boon. Basically, we like to provide power. The more power a chip needs, the better, so that's good. With AI, things like the NVIDIA Grace Hopper chipset, you're now at about 1,600 W per motherboard, which is a huge increase over the typical GPU, CPU. That means that in a cabinet, one rack of server, it used to be about 30 kW per rack. Now, it'll need to be 90 kW or even 100 kW. And so you've got this filing cabinet jammed full of processors, and there's one very thin shelf available for AC to DC conversion. And now you need 3x as much power out of it? That's where the gallium nitride kicks in.
The reference design that we had in Las Vegas last week is a tiny design, but it's 4,500 W, and that's about 4x as much power as the old silicon can do. So very high efficiency, very high power density, and on a dollar per watt basis is lower than the incumbent silicon. So it's a real tipping point, and AI is just dragging that forward. So the GaNSafe technology that we launched, that started shipping last quarter, and that will layer through this year across several platforms at different customers.
You mentioned that GaNSafe started shipping fourth quarter of last year.
Yep.
And obviously, going to ramp this year. Are your initial design wins in the more sort of, I'll call it, standard-based, 3,000 W, you know, kind of data center server power supply with some of these, you know, kind of higher power, much thinner form factors, you know, a little bit further out? Or, or could you see? I mean, I would think that AI, you know, things could move pretty quickly, but obviously, you know, you want to make sure it, you get it right. So.
That's correct.
Don't want to run too fast.
Yeah, the first one is around a 3.2 kW design. And, you know, the great thing about going into that kind of customer is they tend to be the traditional Taiwanese design base. So companies like Delta, Flextronics, Chicony, Lite-On, and they're already our customers for the laptop chargers that then get sold to Dell or HP or other companies. So already on the approved vendor list, we already know them. Navitas is still only 10 years old, but we're not exactly young guys, so we've got the Rolodex and the contacts to really move the designs.
Excellent. And then lastly, I know you sort of started first to ramp in the mobile charger market. I was a little surprised that that was one of your smallest TAMs. But I think some of that's a function of just that these are typically, you know, a year, maybe not even a year-life cycles. Talk a little bit about the trends you're seeing in mobile adapters. You know, where are we in the transition to 75 W, 100 W, 100+ W adapters, and some of the Chinese have made some pretty strong commitments to increasing their use of GaNs, especially in higher end models, Xiaomi, OPPO, maybe talk about some of the commitments they've made.
Sure. Yeah, it's, it's all about speed, you know, speed of charging, speed of adoption, and the speed of the GaN material itself. So GaN is a, is a racehorse. It wants to run at high speed. By doing that, you can shrink the magnetics, shrink the passive elements like capacitors, EMI filters. And what that means is that you can either make a thing really tiny, or for the same size box, pack as much power in as you want to charge things faster. So power gives you faster charging. So, we're now in 10 out of the top 10 mobile OEMs. That has been, in the past, the kind of flagship, just the thin quantity at the top, which is great for advertising, but doesn't, you know, really fill the fab. What we're seeing now is the cascade down from that flagship into mainstream.
This is the Mercedes C-Class retractable seatbelt story. Who in the world needs a retractable seatbelt? Well, everybody's got them now. So OPPO and Xiaomi, you mentioned those names, they've both said that this year, calendar 2024, 30% of all of their charges will be GaN. And what we've seen with the market in general is with 5G, with bigger screens, with streaming video and all of this, power demand is increasing. So Xiaomi, the Xiaomi 14 is a 90 W charger that's in box with Navitas. The Xiaomi 14 Pro is a 120 W charger for a phone in the box with Navitas, and that's a great sweet spot for us. We're far and ahead the the highest performance, and we've got a great price to go with it as well, so it's a really good benefit for us.
We see that trend with all of the phone companies as well. So we'll all be going from, r emember the 6 W or 5 W sugar cube? Well, you'll have something that's the same size, but it'll be 100 W.
As we move up in wattage, how does your ASP and your dollar content benefit? I think you'd said at about 75 W, it, you need the power factor correction, which.
Yep.
A second IC or maybe more than a second IC, but.
Sure, yeah. So, generally speaking, for about a 30 W charger, call it $0.40, $0.45 for one of our chips. So it's quite small, quite a low price. At the 65 W level, it might be $0.80, $0.85, $0.90. And then there's that trip point, that 75 W, and that's a legal requirement to maintain power quality on the grid in Europe, in Japan, and in China. And at that point, you need another chip that just maintains that power integrity. So that means you've got a charge, a chip here that now needs to handle, say, a 100 W and another one here that's a 100 W. Well, now you're talking a $1.50, $2. You go up to 200 W charger, you might be talking $2.50, $3 of content.
As that tailwind of the market needing more power, our cell content goes up as well.
It sounds like a, perhaps a double benefit. As you go up in power, you probably get higher GaN penetration, then as the power goes up, you have higher dollar.
Correct. I would be very surprised if anyone had a cell phone charger that was being designed with silicon today. It's a real tipping point.
Yeah, some of the GaN chargers that you have given out have become, you know, my favorite. So, I'm definitely a convert, so thank you for that.
You're welcome.
I wanted to move to the capacity commitments. You sort of touched on it. I think 3x at TSMC, you know, kind of how does that come online? Is it a take or pay, or do you just have to sort of notify them in advance if you're not going to use the capacity? I know it's been a pretty tight foundry market until probably the last couple quarters, but maybe you can expand a little bit more on those. You know, we'll talk about TSMC when we get to it.
Yeah, I think with TSMC, you know, so first of all, capacity is not an issue there for GaN. I mean, we provide them with rolling forecasts from which they plan. There's no incremental investment needed by us. There's no take or pay contracts. They're a great supplier, and we're a great partner, frankly, with them on the GaN side. So they run GaN in a facility with other products, and as the GaN demand goes up, then they start ramping, bringing on more capacity for GaN and taking down the others. So, there's nothing in the foreseeable future that would suggest that TSMC can't supply us with the GaN we need. Yeah.
And how about X-FAB and the 5x capacity adds that, y ou know, is that all online by the end of this year, or does it kind of tail.
Yeah.
Into 2025?
Yeah, it's coming online and I think X-FAB has made tremendous progress in terms of being able to respond more quickly to our demand requirements. So we're starting to see them catch up. And if you remember, a year ago, let's say, we would talk about how we were supply-challenged with, and the fab being one of them. But they are adding capacity and are on track to get the 5x. And the 5x reference point is going back to 2022, but they are on track to have that in place, and that's happening within the next year or so.
Great. And then lastly, on the capacity front, part of the proceeds from the recent offering was to bring, spend $20 million on epi reactors to bring.
Right.
Some epi, about half of your epi requirements, in-house over time. Just talk about the benefits of in-house epi.
Yeah, sure. So the $20 million investment is for three reactors, the first of which should be up and running middle of this year. And really, it's a huge, frankly, a large cost savings relative to outsourcing the epi process. So that was one benefit. The other is certainty of supply. We've seen that epi as a bottleneck. And those are really the two factors. And three, there is still. It's kind of like the analog world, there's still a little bit of black magic about it. And we've hired a team of epi experts from competitors. So there are multiple advantages for bringing that in-house, including margin expansion.
I was just going to ask you, how do you get the talent to actually run those chambers? Because it's pretty tricky stuff.
Yeah. Yeah, it is. So we steal them. No, we don't steal them. But in all seriousness, we've been very, very successful at recruiting from companies like ON and Infineon and others in the industry. So we've become known, frankly, as from a recruiting standpoint, as, you know, if you want to be in the GaN or silicon carbide space, we're being very successful at recruiting from other companies. Yeah.
Perfect. Just a few more financial questions. You know, to the analyst that you said a target to grow 6x-10x faster than the power semiconductor market. Level set us, and what kind of growth do you expect from the power semiconductor market?
So for.
Market share guy.
Quoting Yole, the French research company, the overall power semiconductor market, which includes silicon, silicon carbide, and GaN, should grow at 9% in 2024.
Okay.
Of which, the bulk is still silicon, and that's growing only at 5%. So silicon carbide and GaN works at about 29% growth forecast. Although that forecast was before the recent kind of inventory checks, you know, news from people like Microchip and people like this. So, that's our baseline for at least for one year. But we see us, because we're a pure play, we don't have the legacy, the incumbent anchor of silicon, as growing incredibly fast compared to the market.
You know, should we think, you know, if I'm thinking, let's stick to a three-year cadence. It's growing high double digits for five years is great job for anyone. But, you know, next three years is something 30%-50% higher. What's it, you know, obviously, you know, 9% is the baseline for 2024. We can do the math, 6x-10x that. But, you know, going through 2025, 2026, you know, do you think you're still in that 30%-50% growth rate range?
We haven't predicted that publicly. We said that this year will be 50% or more, and we've given those long-term estimates, but definitely it's a, it's a growth story. It's all about getting market share and also taking it away from silicon. We've got to remember that the silicon technology is still the old guy that we need to, to defeat with the right technology, at the right price, and the right quality.
But in some ways, that probably makes it easier because the market's already there. You just have to convert it rather than grow.
You do, but as any technology goes from old to new, the old stuff hangs around quite a bit. It doesn't immediately die overnight. Legacy designs, length of design, you know, there's always a laggard who still wants a steam engine.
Gross margin target, 50% +?
Yep.
You're at 42.5% now. What are the, what are the biggest factors to get us from 42.5% to 50%?
Sure. So, so there are really three factors when we, when we talk about gross margins and the long-term model. So, so the first, and we, we touched on it a little bit, is epi, and, and that's really strategic manufacturing investments, right? So, so epi would be the first, there will be, you know, more to come, and it could be, you know, at, at, at the fab level in terms of maybe a joint venture or working with a partner on, on that. It could be in the packaging level, but, but there are more to come, to help drive costs on the cost side. So we're, we're very focused on that.
End markets, diversification, I mean, we touched on pipeline, and we're seeing kind of markets outside the mobile space, and mobile is very big, and we're committed to it, but we're seeing, you know, a lot of opportunities, strong growth in these opportunities outside mobile. So as our end market and our business diversifies into these other markets, those generally have higher margins, so that'll be an uplift to margins. And last is technology, and, you know, the company historically has been able to iterate technologies kind of one generation to the next in a roughly 18-month cadence. Which is frankly faster than any other company I've been at or seen in this industry.
And that also has the benefit of, again, driving the value proposition and, in fact, creating more space between us and the competition, and we're able to charge for that. So really, those three things are ultimately what drive margins higher.
Yep.
Yeah.
Lastly, can you level set us or remind us, what is the rough quarterly break-even run rate and what op margin or gross margin?
Yeah.
Are you assuming, you know, at that quote-unquote break-even rate?
Yeah. Yeah, yeah. So what we've indicated is, you know, from a break-even at the operating line is $50 million-$55 million a quarter in revenue. You know, we guided midpoint last quarter, and it was around $25.5 million, to give you a sense of where we are relative to that. The good news is, you know, we have a really clean balance sheet. We had $170 million in cash at the end of Q3, so, you know, plenty of capital on the balance sheet to fund the business to that. In terms of gross margins and OpEx and what you need to get there, I think it's a little too premature.
I will say we're focused on the growth side, as opposed to cutting spending or something like that to get there. So we'll continue to invest in the business. There is operating leverage in the business, and we're seeing that now. In Q4, again, based on the guide, you know, year-over-year, we said a 100% revenue increase and 20% increase in OpEx. So we're starting to see that leverage, but all things being equal, we will err on the side today on growth. So it's, y ou know, I couldn't give you a, "Here's a gross margin number, here's a OpEx percentage." I think that number will move around a little bit. Yeah.
All right. Well, we are at the end of the time for this session. So, Ron, Steve, thank you very much for joining us at the Needham Conference. We really appreciate it.