Good morning, everybody, and thank you for joining us on the final day of Needham's 27th Annual Growth Conference. My name is Quinn Bolton. I am a semiconductor analyst for Needham & Company. It's my pleasure to host this fireside chat with Vicor Corporation. The company is a leading supplier of high-performance power components and complete power systems based upon a portfolio of patented technologies. Vicor's easy-to-deploy modular power solutions for power delivery networks provide the highest density and efficiency from source to point of load. Joining me from the company are Chairman, President, and CEO Patrizio Vinciarelli, Corporate Vice President, Global Sales and Marketing Phil Davies, and Corporate Vice President, CFO, Treasurer, and Secretary Jim Schmidt. Patrizio, Phil, Jim, thank you for joining us.
Thank you.
I'm going to hand it over to Jim just for the quick safe harbor language, and then we'll get right to Q&A. Jim?
Our remarks that follow, including answers to your questions, may contain statements that we believe to be forward-looking statements within the meaning of the Private Securities Litigation Reform Act. These forward-looking statements are subject to risks that could cause actual results to be materially different from those expressed or implied by such forward-looking statements. These risks include, among others, matters that we have described in our most recent Form 10-K and Form 10-Q, and other filings we make with the SEC. Further, Vicor has not yet disclosed our financial and operating results for our fourth quarter ended December 31, 2024. Therefore, we are unable to discuss or answer questions regarding our financial and operating results or developments for periods after our third quarter ended September 30, 2024.
Perfect. Patrizio, Jim, Phil, I wanted to start off with sort of an update on the ITC litigation against Delta Electronics. You had a positive initial determination back, I think, in October. Can you give us the latest on that case? I think we're expecting a final determination here by early February. But what's the latest on the status of the ITC case against Delta?
We're making good progress. To your point, a final determination is expected in early February. We expect that that final determination will bring about an exclusion order that would, at the outset before the completion of the presidential review period, trigger the need for bonds to be posted on any infringing computer equipment that would be allowed to be imported into the U.S.
Got it. Okay. And on the last earnings conference call, Patrizio, you'd mentioned that Delta is talking to its customers about potential workaround solutions. Have you seen anything from Delta on that workaround solution? And perhaps more importantly, to the extent that Delta is claiming a workaround solution, does Vicor have any input into the customs office to help determine whether or not those new modules actually successfully avoid your patents?
We have seen some evidence of an attempt at a workaround. I think the best way to answer that question is by pointing to the reaction of some of the major customers. And the ones we spoke to, I don't believe this workaround is effective. They believe that the workaround still results in infringement of some of the Vicor patents that have been asserted.
That's what you're hearing from the customers, is that they still believe the workaround may infringe patents?
Yes. The major customers are very skeptical of the effectiveness of the workaround.
Got it. Okay. The other question just sort of related to this lawsuit. Today, Vicor has one OEM licensee of your patents. Have you seen increased interest from other OEMs in licensing your patent portfolio as a result of the initial determination and expected final determination in the ITC case?
Yes, we have.
Okay.
We're making good progress both with OEMs and non-OEMs. So to your point, one notable example is an OEM supplier of AI hardware. We've also made progress with respect to non-OEMs, hyperscalers that employ otherwise infringing hardware in data centers.
That sounds like very encouraging progress, Patrizio.
I think we're doing quite well. We're doing well at the ITC. We're also doing very well with respect to retaliatory actions that Delta has brought in this court. We are making great headway with respect to those three lawsuits. So I'm very satisfied with the progress that our legal team has achieved over the last year.
Well, we certainly look forward to potential additional announcements on the OEM licensing front. I did want to ask, your current licensee has seen a pretty impressive growth in the royalty revenue stream just from that single licensee, if I remember. Royalties back in the first quarter of 2023 were about $2 million. I think in the September quarter, royalties up to $13 million plus. What's your outlook for continued growth in the royalty business from the existing licensee? And I guess a related question, is the royalty typically paid on the amount of power converted in a module, or is it simply a function of the number of units that are imported into the U.S. by this existing licensee?
So the answer is that it depends. If it is an OEM license, then to your point, it's proportional to the aggregate power of all the modules that are licensed. If it is a non-OEM license for captive use within data centers, it is based on the anticipated usage within a two-year timeframe.
Okay, so different models, but I guess on the existing OEM license, it is related to the power consumed, and so as an example, a 1,000-watt power module would have a higher royalty than, say, a 500 watt module.
Yeah, it would have double the royalty.
Perfect.
Everything else being the same. There are many factors at play, but power for an OEM license is the primary factor.
Perfect. I wanted to move to your Gen 5 technology and vertical power delivery. For investors that may be less familiar with your Gen 5 technology, can you provide a brief overview of the advantages of Gen 5 architecture over the Gen 4 architecture?
So Gen 5 provides much higher current density. And current density is key to enabling high-performance solutions. So generally speaking, with our earlier generation, we have added capability in the 1.5 ampere per square millimeter range. With Gen 5, we've raised the bar to the 4.5 ampere per square millimeter level. That's a key differentiator. It's a key differentiator in that high current density is needed, maybe at this point in time, not at the level of 4.5 ampere per square millimeter. But having that current density makes for a much simpler implementation of the system because our modules, unlike VRs, don't need to take up all of the space that is available on the backside of GPUs, TPUs, high-density processors. And that facilitates a much simpler mechanical thermal management of the solution.
It sounds like the Gen 5 with that significantly higher power density really enables a vertical power delivery architecture rather than the lateral power architecture used by most XPU vendors in the market today.
It really facilitates both. Laterally, it can get closer. But ultimately, as AI systems continue to advance their capabilities, there is, as far as the eye can see, increased demand for greater and greater current density.
As you think about vertical power delivery, is there a threshold for power or current that as these XPU vendors get above that threshold, they really are going to need to go to vertical power architectures?
I don't know that there is a threshold that is set in stone. I think it depends, generally speaking, on a variety of factors with respect to the system and the solution. But one way to look at it that makes fundamentally the improvement and the capability clear is that at the level of 1,000 amps, as a ballpark in a GPU, the savings in terms of reduced power dissipation in the so-called power distribution network, or PDN, is of the order of 10%. Not exactly 10%. It could be 8%. It could be more than 10%. But needless to say, the 10% ballpark number needs to be viewed in terms of its impact on total power consumption. As we all know, data centers, their deployment is very much power delivery constrained, in addition to, obviously, the cost of the utility bill.
But there is fundamentally a bottleneck with respect to power delivery. And that is one reason why VPD is becoming a necessity in order to improve efficiency, be able to deliver more compute power, make the cost of that compute power substantially lower. Not to mention enabling aspects having to do with the fact that high efficiency means reduced heat generation, which has implications with respect to mechanical and thermal management of the solution.
Yeah. I was just going to say, I think half a point to a point of better efficiency is significant. So eight to 10 certainly sounds like that is a very meaningful difference. Do you have a sense over the next couple of years whether you would expect pretty broad adoption of vertical power delivery, or is it still sort of somewhat unclear when we'll see mainstream adoption of vertical power?
It is a necessity. As of now, the implementations of VPD are what we call first-gen VPD, which is fundamentally another Vicor innovation, Vicor SiP, with respect to its first-generation VPD. We're not deploying it because we've invented a more advanced form of VPD enabled by Gen 5 higher current density multipliers, which greatly facilitates the scalability, cost-effectiveness, robustness of the solution. But we are aware of at least two major companies that have deployed with mixed success, particularly when it comes to manufacturability and robustness, first-generation VPD.
Okay. Speaking of your Gen 5 technology, can you give us an update on adoption or when you expect to see the Gen 5 technology ramp? I believe you're working with a lead customer that will be first to ramp. And then have you started broader sampling of the Gen 5 technology to a larger customer base at this point?
So we're still primarily focused on our lead customer. The system solution is comprehensive in terms of multiplicity of components to the power system solution with 5G building blocks in a VPD, second-gen VPD type application. So we're primarily focused still on a lead customer, which will ramp this year. We have not yet made demonstrations to other customers or potential customers. There's great interest with respect to our solution on a number of fronts. But we want to make sure that we have completed all that is necessary to ensure a robust, scalable, high-performance solution. We're very close to that milestone. We're not quite there yet. And we're keeping our powder dry with respect to other engagements until we get there.
Okay. And from sort of sampling of Gen 5, once you start that process, how should investors be thinking about the timeline from those initial samples to when the Gen 5 could ramp in production? Is that typically about a 12-month design-in cycle and test validation process, or is it shorter? Is it longer?
Starting with our lead customer, the lead time to production is on the scale of six to nine months, so later this year. With respect to other customers, I think the timescale will probably be closer to a year because, in a way, they're further behind with respect to engagement.
Got it. Okay. And then one of the drawbacks of sort of the vertical power architecture, maybe cooling since these components sit underneath or on the backside of the processor, does Vicor's architecture allow for adequate cooling of the MCMs underneath the processor?
So again, we have two architectures. We have our first-generation VPD, which is a stacked form of power delivery, meaning there's a multiplicity of layers, including a capacity bypass layer and multi-cell converter layer. And to your point, with that stacked approach, thermal management is a challenge because heat needs to traverse the various layers of the stack. And each layer is an impediment to the heat path, right? And the accumulation of these layers results in a thick solution, typically in the 4-5 millimeter thickness range. So heat has a very difficult time getting through that, through the individual layers and the stack as a whole. With our second-generation VPD enabled by Gen 5 components, with much higher current density, the complete solution is only 1.5 millimeter thin. It's got very, very low thermal resistance per unit area.
So it actually can contribute to the cooling of the GPU or TPU or XPU by virtue of its very low thermal resistance.
Okay. So, lower thermals, and then again, maybe just in comparison to sort of a vertical power architecture for multi-phase, can you just sort of discuss the number of components that Vicor would supply for vertical power versus what your multi-phase competitors are looking at?
So it's not atypical for a multi-phase solution in the thousands of amps to require upwards of 50 VRs. Again, these things, as part of their first-generation VPD, are stacked on top of capacitor layers. The stack as a whole is quite thick and thermally inefficient, very challenging. With our Gen 5 VPD, we are in a much better position in that we can take, let's say, a multiplicity of 50 VRs down to one or two current multipliers, depending on current level and also depending on the architecture of the PDN. So there is a major reduction in the number of components in the X and Y. And there is the elimination in the Z of the stacking requirement that is at the heart of first-gen VPD.
Okay. We had a question from the audience, and this may be a good sort of place to sort of fit it in. The question is, a leading GPU manufacturer is having heat problems. Does Vicor have everything needed to satisfy that customer's needs and objectives?
We believe we do.
Okay. Yeah. Well, obviously, not looking for you to pre-announce customer relationships, but certainly sounds like, especially with Gen 5, that the heat and current density would be pretty significant advantages for Vicor over the multi-phase solution.
Yeah.
Perfect. Wanted to switch over and talk about some of your other end markets, starting with the automotive opportunity. Can you give us an update, perhaps, Phil, where you guys are with the number of OEM engagements and sort of design win momentum in the automotive market?
Sure. Yeah. The momentum continues in a really nice positive direction. I think on some of the previous calls, I've mentioned that we have two OEMs that we are going into production with at the end of this year into 2026. Our focus is obviously on adding additional OEMs to that list and also Tier 1 engagements as well. Our model there is to really we've been focusing heavily on OEM pull, if you like, getting the OEMs excited about the density of the solutions, the low weight that we offer, the extended range that we offer to them. And then that creates a pull onto the Tier 1 infrastructure that typically builds the systems. And so we've seen a higher level of engagement as a result of that strategy with Tier 1s.
And so at this point in time, we have two OEMs, as I said, ready to go into production. We have, with one of those OEMs, a number of new platforms that have been added to the initial design-in. So there are three or four ranges of vehicles that we are designed into now. And then we have three additional OEMs that are moving into what we call the design-in phase. We're moving out of collaboration opportunity with building systems and moving into design-ins that should ramp to production then in 2027. So we're heavily focused on those three OEMs, and they have a directed supplier model. So they bring in a Tier 1 for us to work with.
Then we have numerous collaborations and opportunities with additional OEMs and Tier 1s on onboard charger systems, 800-volt to 48-volt zonal architecture, DC-DC converters, 48- to 12-volt zonal architecture converters as well. So it's really a mix of different systems. And I'm really excited still by that opportunity that lies before us. But as I have mentioned, it's a couple of years behind where we thought it would be due to COVID and other factors. But we're looking at 2028 and 2029 for real ramps of the additional OEMs in terms of our business plan.
And yeah, I was going to say that was one of my questions, just kind of talk about the slope of this ramp in the automotive business over the next several years. I think the first two OEMs you've described back at the June shareholder meeting as sort of being higher-end or premium vehicles, and the more mainstream, higher volume programs would probably kick in a couple of years later. But maybe just spend a minute, just how investors should be thinking about the slope of that ramp in the automotive business, say, through 2028, 2029. When do you hit the knee in the curve, I guess, is the gist of that question?
I think the knee comes really in 2028 with the addition of the three additional OEMs because they typically have higher volume platforms. They're not so premium. They have premium platforms, but not in that real high end of the market. So the applications that we're going into initially in 2027 will be things like active suspension type programs where the premium vehicles that they offer have that capability. And so some of the early designs will be in lower volume premium vehicles where we will prove ourselves. That's typically what automotive companies will do. They give you proving platforms and proving applications. And then having proved ourselves, we will then move on to higher volume systems, DC-DC converters, and sort of mainstream applications powering the vehicle.
Okay. We talked about the new Gen 5 technology sort of for the high-performance computing market. Does the Gen 5 technology also provide similar advantages for your automotive customers?
Yes, they do. Particularly on the DC-DC lower voltage, 48 volt regulator side of things. We're going to be taking our regulator platform for high-performance compute, the Gen 5 PRMs and a new buck regulator that we've been developing. And we're going to be bringing that through into automotive applications, either for 48 to 48, stiffening of the 48 volt systems, zonal architecture, or doing 48 to 12 for legacy loads. And we're having conversations right now with OEMs on the Gen 5 regulator family that we'll be bringing through next year, sorry, later this year and into next year.
You've mentioned sort of in this discussion a number of products for automotive. Is there one or two where you're seeing the greatest interest, whether it's kind of 800, 400 volt conversion from the battery to the 48 load or for onboard charging or roadside chargers? Is it the 48 volt regulator? Is it 48 to 12 step down? Or are you seeing pretty good interest across your automotive portfolio?
It's across the portfolio, but the first two design-ins are 800 volt to 48 volt DC-DC conversion. These new opportunities that I talked about with the three additional OEMs are on high voltage to 48 volt conversion, 800 to 48, that type of thing, 400 to 48. And then following on from that, the 48 to 12 regulators I spoke about, the non-isolated and non-regulated products that we have, 48 to 12 as well. So it's quite a mix as you go across all of the opportunities that we have.
Perfect. Wanted to shift now to sort of your outlook for the aerospace and defense and industrial businesses. First off, what are you seeing in terms of inventory? Is there any excess inventory in those markets, probably more on the industrial side, perhaps than aerospace and defense? But the industry's obviously been through a pretty tough patch for the last one to two years, as many end markets have digested inventory. Is that something you see in the industrial side of your business?
I would think that we haven't really seen a lot of inventory in our channels or at our customers on inventory for our products across the old brick products, if you like, and some of the advanced products. We did see some inventory buildup in China, some of our Chinese distributors, but that now is starting to work its way through, and we're pretty much done with that, but I think that was a select area where we did see some inventory buildup in the last couple of years, but that's worked its way through now.
Okay. And just for folks listening to the webcast, I just want to remind everybody, if you do have questions, please do submit those through the dialog box at the bottom of the screen. I will try to work them into the discussion, but just a reminder. Phil, coming back to the sort of aerospace and defense and industrial market outlooks, can you remind investors sort of what are your TAM expectations and sort of growth rate expectations for those two end markets?
Yeah. So as I mentioned at the shareholders meeting on the last call, our goal is to double our industrial business in the next five years and double the aerospace and defense business in the next seven years. As I've spoken about, we have what we call the top 100 customers that we are focused on globally. Those top 100 customers provide us the opportunity to get to our North Star revenue goal of $1 billion. I mean, they have the SAMs, the growth, the positions in the markets to be able to do that. And on the aerospace, defense, and industrial side, we've selected there's about 50 to 60 customers there that we're laser-focused on, really good design-ins with. And the new products that we're bringing through now will add to the growth of some of the products that we introduced three, four, five years ago.
That's the nature of those two markets. So I'm very encouraged by the position that we have in the top 50 or so of the top 100 in those markets. And the new products this year will add to that. And I think that out of the $1 billion, we're looking for that to represent $500 million-$600 million out of those two market segments.
Great. Wanted to move to some financial or sort of business-related questions. First, in the past, Vicor has been affected by tariffs on the import of semiconductors and other components from China. Obviously, the incoming administration has made a lot of noise about increasing tariffs on Chinese imports. To the extent that those are passed, what do you think the impact on the business would be? Have you been able to source components from other geographies over the past several years to maybe reduce that tariff risk to the business?
Yeah. So, Quinn, we've really substantially reduced tariffs over the last couple of years. And the way we've done that is we set a goal to do that with a supply team and a purchasing team that really has gotten very good at moving production. So we have de-risked supply chain. We've moved supply chain out of China. We've reduced tariffs by a lot, a very substantial amount. Now, in terms of going forward, the environment for tariffs, we have a variety of external sources that are giving us all kinds of guidance and opinions and white papers. And so we've studied it. But I think you know that the reality is no one can be completely certain what that environment's going to look like. What I would say is that we now have muscle memory in the company around being able to respond to this.
And we've got a successful approach, and the team is very effective. One other point I would make is that it's important to remember that Vicor is one of the few remaining semiconductor companies in the U.S. that manufactures in the U.S., and pretty much exclusively in the U.S. And what I think we're seeing is that's becoming an increasingly important competitive advantage for Vicor. So I think there's two sides to the tariff story to consider here.
No, I understand. I think you're right. The local manufacturing does seem like it's becoming increasingly important to a number of industries. So thank you for pointing that out. Maybe for Phil, not trying to be too near-term here, but book-to-bill is hovered about parity for most of 2024. Can you share any thoughts on when you think the company might get back to a sustainably positive book-to-bill above one-to-one?
I think that we're looking at 2025 being a very good year for Vicor. I think Q3, we got over one in terms of our Book-to-Bill. I mean, our objective is to maintain that. I think what we're looking at right now is encouraging. And so I think we're going to have a good 2025.
Excellent.
Fiscal year, that 2024, was a transition year and transition is over, and 2025 is here, so we look forward to 2025 being a record year for Vicor.
Excellent. Excellent. And obviously, we'll stay tuned for the earnings results here over the next few weeks. We have had another question come in from the audience. I'll read that now. What do you think has kept the largest GPU vendor from using vertical power delivery as aggressively as they should, given the enormous benefits of VPD? Are there some technical challenges or perhaps diminishing benefit for vertical power delivery in future gens because of liquid cooling or other technologies?
What has prevented widespread or what you might call mainstream adoption is the kinds of challenges that I was describing earlier. The fact that first-generation VPD, as conceived by Vicor as a first-generation technology, has the stacking challenges. So leading OEMs, rightly so, came to the conclusion that the technology was not scalable, cost-effective, robust enough, mature enough for wide-range deployment. And particularly OEMs with a leading position that fundamentally can sell everything they can make and have very little competition at this point in time, rightly so, make a judicious choice of availability and scalability at some compromise in performance as opposed to achieving the highest performance they could with the right technology. And that's where second-gen VPD comes into play.
Thank you. I wanted to ask about the Andover expansion. Is that now effectively complete? If not, what work is left to complete, and how much CapEx do you expect to spend on that facility in 2025?
We've internalized most all the manufacturing, Quinn. We really have a better profile for absorption, etc. Out-of-pocket spend has dropped. We're really a significant. I don't know the percentage. Maybe Patrizio would quote a number. But we're there in terms of the in-house capacity. We can add another big increment to the plating line when we need to. CapEx this year, we expect to be in the range of $20million-$30 million.
Okay, so fairly modest levels of CapEx going forward.
Yeah. We have quite a bit of capacity.
Okay. And then can you just discuss how customer qualifications of the expanded Andover facility are going? Have you now completed those customer qualifications with most of your large customers?
So we're doing well. The integrated capacity has capabilities beyond what we were able to achieve in the past with outsource process steps. Take as an example the NBM, which is obviously a key component in many respects, right? With our earlier generation capability, at the package level, we could only support an MSL4 capability, which requires that these components be properly stored in an environment that is not humid. But with our vertically integrated process capability, we've been able to raise the bar so that our new NBMs are MSL3, which facilitates their deployment on factory floors without nearly as much care with respect to exposure to humidity. And that's a tangible measure of the process improvements and benefits of vertical integration.
Great. Lastly, for me, just can you spend a minute discussing your outlook for the brick products business in 2025? I know this isn't quite as exciting as advanced products, but it's still a meaningful part of the business. Do you think that business can grow in 2025, or would you expect it to be perhaps more flat year on year?
I think bricks, being nearly 40 years old as a power component paradigm, will do well, essentially stay level, right, and continue to live for a few more decades. We should not be looking at them as the growth engine for the company, right? So the growth comes with our advanced products hardware business and our advanced products licensing business.
Understood. We've got maybe a minute or two left. And so I maybe just hand that minute back to you, Patrizio, Phil, and Jim. If there are any closing comments or thoughts you'd like to just leave with investors listening to the webcast today?
I would just go back to saying that, as Patrizio mentioned, and I talked about a little bit, we're looking forward to this year. We've made a lot of progress. I think it's going to be an exciting year for the company with the Gen 5 technology to be rolled out and many new products across aerospace and defense and industrial, and I really am excited about the top 100 focus that we've brought to the company and the opportunity that those customers that are located globally offer to us with the technology that we offer. They really value the value proposition, which is density, current density, power density, low-weight system solutions, modular building blocks. I'm really excited about what we've done in terms of that, honing the focus of the company to those top 100 accounts.
Great. Well, Patrizio, Phil, Jim, thank you very much for joining us again at the Needham Growth Conference. It's always a pleasure to host the Fireside Chats with the company. And certainly hope we can do it again in future years. And look forward to catching up with you guys all after earnings.
Thank you.
Sounds good. Thank you. Bye-bye.
Thanks, everybody.