Welcome everyone to the Vicor earnings results for the third quarter ended September 30th, 2019 conference call hosted by Dr. Patrizio Vinciarelli, CEO, and James Simms, CFO. During the presentation, the lines will remain on listen only. If you require assistance at any time, please key star zero on your phone and a coordinator will be happy to assist you. I'd like to remind all parties that this conference call is being recorded for replay purposes. With that, firstly, I would like to hand the call over to Jamie. Please go ahead, sir.
Thank you, Chris. Good afternoon, everyone, and welcome to Vicor Corporation's earnings call for the third quarter ended September 30th, 2019. As stated, I'm Jamie Simms, Chief Financial Officer, and with me here in Andover is Patrizio Vinciarelli, Chief Executive Officer. After the markets closed today, we issued a press release summarizing our financial results for the three months and nine months ended September 30th. This press release has been posted on the investor relations page of our website, www.vicorpower.com. We also filed an 8-K today related to the issuance of this press release. I remind listeners this conference call is being recorded and is the copyrighted property of Vicor Corporation. I also remind you various remarks we may make during this call may constitute forward-looking statements for purposes of the safe harbor provisions under the Private Securities Litigation Reform Act of 1995.
Except for historical information contained in this call, the matters discussed on this call, including any statements regarding current and planned products, current and potential customers, potential market opportunities, expected events and announcements, planned capacity expansion, as well as management's expectations for sales growth, spending, and profitability are forward-looking statements involving risks and uncertainties. In light of these risks and uncertainties, we can offer no assurance that any forward-looking statement will in fact prove to be correct. Actual results may differ materially from those explicitly set forth or implied by any of our remarks today. The risks and uncertainties we face are discussed in Item 1A of our 2018 Form 10-K, which we filed with the SEC on February 28th, 2019. Please note the information provided during this conference call is accurate only as of today, Thursday, October 17th, 2019.
Vicor undertakes no obligation to update any statements, including forward-looking statements made during this call, and you should not rely upon such statements after the conclusion of this call. A replay of the call will be available beginning at midnight tonight through November 1st, 2019. The replay dial-in number is 888-286-8010, followed by the passcode 70495083. This dial-in and passcode are also set forth in today's press release. In addition, a webcast replay of today's call along with a transcript will be available shortly on the investor relations page of our website. I'll start this afternoon's discussion with a review of our financial performance for the third quarter and nine months year to date, and Patrizio will follow with his remarks after which we will take your questions.
Beginning with consolidated results as stated in today's press release, Vicor recorded total revenue for the third quarter of $70.8 million, up 11.7% from second quarter revenue of $63.3 million, but 9.3% lower than our record revenue of $78 million recorded for the third quarter of 2018. Year-to-date revenue stands right at $200 million, off from 2018's nine-month total of $217.5 million, reflecting the trough in data center shipments of the first half of this year. Third quarter revenue growth reflected a resumption of shipments to contract manufacturers on behalf of an existing data center customer. Production is resuming after the pause of the past two quarters. We also ship pre-production volumes of advanced products for customers to use in their own pre-production activities.
Quarterly revenue from brick products was unchanged sequentially, as double-digit domestic growth was offset by tariff and trade-related declines in China and Hong Kong, as well as ongoing economic weakness in Europe. For the third quarter, the brick advanced revenue split was 66% brick, 34% advanced, in contrast to 76% brick, 24% advanced for the second quarter. International revenue rose 27.6% sequentially, reflecting the aforementioned increase in advanced products shipped to offshore contract manufacturers, offset by a slight decline in exports of brick products. International revenue rose to 58% of revenue for Q3 from 51% for Q2. Consolidated gross margin as a percentage of revenue was steady sequentially, rising slightly to 46.6% from the prior quarter's 46%. The positive impact on absorption of the quarter's increased production volume was offset by higher tariff charges, increased reserves, and mix.
For the year-to-date period, our consolidated gross profit margin stands coincidentally at 46.6%, compared to 48.3% for the 2018 nine-month period, with the lower figure primarily attributable to lower absorption and higher tariffs. We discussed last quarter our development of a duty drawback program, but we now believe any recovery of tariffs paid will occur in 2020 given the backlog of applications being processed by the U.S. Customs authorities. We also continue to pursue opportunities to use suppliers that are not subject to Section 301 import tariffs and anticipate completion of certain important vendor changes in 2020. Quarterly operating expenses were flat both sequentially and on a year-to-date basis. Non-cash stock compensation expense totaled $753,000 for the quarter, in line with prior quarters. Quarterly operating income rose to $6.1 million from the prior quarter's $2.4 million, representing operating margins of 8.6% and 3.8%, respectively.
Year-to-date operating margin stands at 6.5%, in contrast to the 2018 year-to-date figure of 11.7%. Our year-to-date effective tax rate stands at 5.9%, down from 7.3% as of June 30th. The effective tax rate for the third quarter on a standalone basis was 4.3%, essentially unchanged from the rate similarly calculated for the second quarter. Net income attributable to Vicor totaled $5.9 million for the third quarter, representing a fully diluted GAAP EPS figure of $0.14. This is in contrast to Q2 2019 net income of $2.6 million, representing fully diluted GAAP EPS of $0.06. Our fully diluted share count for the Q3 EPS calculation was 42,194,000, which is the sum of both common share classes, consisting of approximately 28.5 million registered and listed common shares, approximately 11.8 million Class B common shares, which are neither registered nor listed, and approximately 1.9 million diluted stock options.
Turning to our balance sheet, cash and cash equivalents sequentially rose by $9.7 million to $81.2 million, primarily due to a positive shift in working capital. Trade receivables net of reserves totaled $39.8 million at quarter end, up sequentially 4%, with DSOs at 44 days, a slight improvement sequentially. All balances are current. Inventories net of reserve decreased 9% sequentially to $49.7 million, with declines in raw materials, WIP, and finished goods. A portion of the decline was associated with the shipment of finished goods originally scheduled to be shipped last quarter, but other reductions are associated with improved planning and improved supply chain conditions. We hope to achieve further efficiencies in inventory management, improving our annualized turns metric, which today stands at three. Capital expenditures for Q3 totaled $3.3 million, an increase of 28% sequentially.
Turning to headcount, our employee total as of September 30th stood at 1,023, up four from the prior quarter's 1,019. Regarding our capacity expansion, we expect to complete the previously announced purchase of land adjacent to our Andover facility during the current fourth quarter. Construction over the winter months will be limited to activities such as excavation, grading, paving, and the like. We expect to begin construction of the new wing of our Federal Street factory in the spring and to complete occupancy in 2020. As stated before, we anticipate internally funding both the construction and the planned two phases of equipment installation. Turning to the fourth quarter, we are positive about our near-term prospects in AI acceleration and supercomputing, but cautious regarding the timing of a broader recovery of data center spending.
As previously disclosed, we have a multiplicity of design wins and expect orders in Q4 for production volumes to be delivered across the first half of 2020. With that, I'll turn the call over to Patrizio.
As Jamie suggested, current booking patterns continue to reflect U.S.-China trade tariffs, macro uncertainty, and weak demand visibility. Bookings for the third quarter totaled slightly over $60 million, essentially unchanged from the prior quarter. Fourth quarter bookings are forecasted to increase by about 30%. Advanced products rose in bookings 40% sequentially in Q3, driven by the resumption of server production by a major customer. We expect further substantial orders from these customers. We also received orders for initial builds of next-generation GPUs from another major customer. Additional orders are expected as this customer accelerates rollout of its next-generation platform. As Jamie mentioned, we also expect orders associated with other design wins beginning in Q4. However, given lead times, shipments from these orders will not occur until 2020. We're well-positioned with leading innovators in artificial intelligence, offering power system solutions that enable advances in computing power.
We expect these innovations will provide growth beginning in 2020. Turning to automotive, current booking activity includes notable design wins from automotive OEMs planning to use 48-volt power system solutions to improve the performance of mild hybrid and full EV vehicles. Autonomous driving systems are also converting to 48-volt to power sub one volt ASICs with current levels approaching 1,000 amps to enable the compute power and intelligence necessary to make autonomous driving safe. As these orders involve substantial investment in our proprietary solutions through NRE payments, we see them as evidence of a long-term commitment to Vicor and to our highly differentiated solutions. As with AI and supercomputing, automotive power system architects are turning to Vicor for a matched performance that in turn differentiates the performance of their end products.
To conclude prepared remarks, having faced challenging conditions since about this time last year, Vicor has uniquely positioned itself for long-term growth and profitability by penetrating key markets with clearly superior and enabling power system solutions. We'll now take your questions. Operator?
Absolutely. Everyone, your question and answer session will now begin. If you wish to ask a question, please press star then one. If you would decide to withdraw your question, simply press star then two. Thank you. We have our first question coming from the line of John. Please go ahead. Your line is open.
Hi, good afternoon, Patrizio and Jamie. Thank you for taking my question. Nice to see things get back on track a little bit this quarter.
Thank you.
My first question is on the orders and bookings that you've seen. Do they indicate that your fourth quarter will be up to what you previously expected, generating some growth, with some new designs ramping? Is that pushed out a little bit into 2020?
For the fourth quarter, we are willing at this point to provide guidance with respect to bookings, which I suggested in my prepared remarks. We expect to be up by about 30%, but we do not want to give guidance with respect to revenues, because the actual revenue number will be affected to a significant degree by terms of business within the quarter, which is not that easy to predict. Last quarter, we made or provided guidance with respect to revenue, which we have lived up to. This quarter, we're providing guidance with respect to bookings, but it's harder to do that with respect to revenues.
Okay. Understood. For the large data center customers that resumed taking your product this quarter, was that because of their spending patterns, something to do with the contract manufacture transition out of China, or some other factors such as CPU shortages being relieved?
I think all of the above in, I think, some varying combination as a function of time. Everything you mentioned has been a factor. Clearly, if we go back to this time last year, given the climate of the industry at that point in time, I think there was a lot of demand that may have gotten ahead of what was needed, in part because of serious shortages, very long lead times with respect to key components. I think many companies got ahead in terms of their orders. We saw a contraction that was fundamentally driven by an agonizing reappraisal of what they actually needed over the next several quarters, some inventory that had to be absorbed.
More recently, to your point, there have been other factors at play, including relocating production lines from China to Taiwan, and in some instances, lack of availability of certain processors that got delayed or had issues that caused delays.
Okay, great. You mentioned a major GPU customer ordering more product. Is that related to orders that you took last year, or is this for a new one? What happens to those old orders now that if it is a new generation coming out?
There's a new generation coming up, and we started to receive significant orders for that new generation. Based on the forecast, we expect that to start ramping over the next few months. That's all I can say at this point in time.
Okay. Last one from me. Jamie, you mentioned some tariff, I guess, refunds or clawbacks. What was the amount that you may have expected in the quarter? Was it material?
Well, we didn't expect it this quarter, but we are anticipating that when in fact we get through the queue that we have to deal with, because there are many companies that are seeking similar refunds, that we're going to have a multimillion-dollar refund coming.
That's our expectation. Okay, great. I'm sorry, one more last one. Can you talk about the status of any manufacturing partnerships you may be pursuing to secure future growth?
We're having discussions. As I mentioned in prior calls, we're taking a very long-term look with respect to this. We want to work with the right licensee under the right conditions. We'll let you know once something comes to fruition, but I don't have anything to say today about it.
Okay, great. Thank you so much.
You're welcome.
Thank you, John. Next question is coming from the line of Jeff. Please go ahead. Your line is open.
Good day, gentlemen. James, I had a question. You talked about the OpEx kind of being flat. It was sequentially and even proved as a percentage of sales, quite substantially sequentially with Q3 over Q2, but was up about 300 basis points roughly, year-over-year. What is it going to take to really start moving this OpEx line down to the lower 30, given now we're starting to see what sounds like a re-ramp of new products that are going to add to the top line? What's going to be the contributor there? Are you going to be able to pull back investment now that you've started winning these new products? Or how are we going to start seeing more fall to the bottom line, given the investments the company's made in the Power-on-Package and advanced product line?
Let me take that. As discussed in the past, the bottom line, in our case, will leverage economies of scale with respect to, first of all, production rates and the top line, with operating expenses, in particular, R&D, which is obviously a significant element of OpEx, not scaling in the proportion of revenue growth, but far below that, because we believe that with the infrastructure and fixed costs in R&D, we are in a good position to not only support all of the advancement that we've accomplished, but also to continue to raise the bar with respect to performance of Power-on-Package products, front-end products, the whole power system infrastructure is necessary for future players in the power system industry to succeed. We believe we can accomplish that without significant growth in operating expense levels in terms of dollars.
As a percentage, as the revenues grow, those percentages will come down. That is an element that will contribute to your point in terms of reducing OpEx as a % of revenue, contributing to the bottom line. Most significantly, on the gross margin front, as we discussed in the past, there, too, we have a cost model that has got a very large component that is fixed. We believe that with increased production rates, we're going to see better gross margins, the bottom line will be driven by expanding gross margins and reduce operating expenses as a % of revenue. Some of similar comments apply to other operating expenses that will not scale in proportion to the top line.
Patrizio, thank you for that answer. I was just wondering, what are the assumptions, given the fixed cost nature of the improvements in gross margin, what's your utilization factor at those levels in terms of where we need to be at utilization levels to start kind of seeing that improvement on the leverage of the fixed cost in the gross margin?
There is no magic threshold. I think fundamentally, incremental revenues with incremental production rates contribute to the margins and contributes to the bottom line. I think what we've been seeing of late, as suggested in Jamie's preparing remarks, in terms of margin in the neighborhood of 46%-47%. We've talked a moment ago about the fact that we expect to be able to pull back some of those tariffs. Obviously, those elements are not reflected in what we're reporting for costs because we have not gotten that money back. Going forward, again, we expect to see, because of initiatives we've taken with respect to alternate sources outside of China, and so on and so forth, we expect to see the impact of tariffs mitigated. More significantly, the impact of increased production rates contributing to margins rising above 50% and more beyond that.
Okay. Thank you. I'll just try to get a couple of quick ones in here. Where are we in the process of moving some of the supply chain into Taiwan? Where is that? Just two quick ones to follow up, and I'll step off here. How much of the backlog right now is represented by advanced products? Has there been any impact to you or any of your customers due to M&A activity that they may be engaged in at this point in time? I'll leave it there. Thank you very much for your time.
Okay. Let me see if I can remember all those questions. I think the first one had to do with the supply chain or moving some of our supply chain from China to Taiwan. To be clear, that comment with respect to changes in supply chain, that had to do with the contract manufacturers that we ship to. In other words, our customers, as opposed to our suppliers, even though in some instances, there have been relocations outside of China already taking place. With respect to M&A activity, I'm not aware of anything. Jamie, are you?
Yeah. Jeff, what did you have in mind?
Well, I don't want to make you guys come out and make comments about customers. Research kind of suggests that someone you may be or may not be working with is in an acquisition now that may or may not go through. I think in my technology background in the cloud and in accelerators and things that you're working on, there could be some hiccups there if things don't go through as planned. The follow-up I had on that, if you can even answer that, is just what's the advanced products in terms of, is it substantially the largest portion of the bookings?
Let me take the one about companies in the accelerator business, and potential mergers or acquisitions taking place in that space. Particularly when it comes to accelerators, from my visibility perspective, we are essentially playing with everybody that has got advanced products in that space. To that extent, it doesn't really matter to us if some of these companies, particularly the startup companies, end up getting acquired because they may change their name, affiliation, but we'll still be doing that business with the acquiring company, that if it is in that space, we're already doing business with or going to be doing business in the next year or two. With respect to accelerators, we are in a very significant, I would say, potentially dominant position given the unique attributes of our technology and the technical needs in that space. Cloud computing is a different story.
We have limited penetration to date in that space. Without question, the advent of 48 volt, which was resisted by some in that space for a long time, and it's now happening with standards that are evolving towards 48 volt, will confront potential customers that have been reluctant to commit to us for a variety of reasons, concerns about single source, Vicor being a very small company. It will confront them with a dilemma, which is, do they want to have solutions that are truly enabling or be handicapped with respect to their products, or accept that they can do business with us, Vicor, having had the history of being a very reliable supplier. I think they're going to go for the latter. Obviously, that has got to play itself out.
Certainly, the convergence to 48 volt in cloud computing, in the data center space, and most significantly in automotive, bodes well for our prospects in that general space.
Okay. Just two quick ones for you, James, sorry. What's been put in place since the inventory correction that occurred at the end of the year into the new year to better understand the visibility you have to the end customer and what their current demand is? Just coming back again on the bookings, is it just the vast majority is coming from advanced products? Just trying to understand the mix of bookings a little bit. Thank you so much for your time.
Unfortunately, when our customers cannot predict their level of business, I don't think we can realistically predict it for them, right? These forecasts, particularly to your point, at the end of last year, suddenly changed. There was some positive feedback, some contraction that brought about further contraction as the mood very quickly changed, I think, starting in August, September of last year. Can we get better at this? I imagine we could, but realistically, I don't think it would be reasonable to expect us to do that much better in that regard. I mean, in our space. These kinds of things happen. They happen periodically from time to time. The mood, if I go back to July or August of last year, was extremely buoyant.
Everybody was worrying about being able to get their components, their power semiconductors, and other components. That mood changed very quickly, very dramatically. I don't know that we could have predicted that. I'm afraid that we haven't made substantial progress with respect to the ability to forecast these kinds of trends in the market.
Thank you.
You're welcome.
The next question is coming from the line of Quinn. Please go ahead. Your line is open now.
Hi, Patrizio. Hi, James. Wanted to follow up just on this new generation of GPU that you mentioned. I think the first generation you won with that customer seemed to be for a fairly limited application. I'm just wondering if you feel that the second generation GPU design that you've won could target a broader set of applications at that customer.
Without getting too specific, I would say that we do agree that the first generation GPU, for a variety of reasons, did not achieve the level of market penetration that the customer had anticipated. I think we have reason to believe, both based on representations of the customer and other inputs, that this second time around, the expectations are most well-founded and likely to come through. We'll take a wait and see attitude with respect to that as well, because again, I suggest in answer to the earlier question, our visibility is largely dependent on the customer visibility. What we understand is that this time around, steps have been taken to ensure much greater penetration, and we'll see what happens.
Thank you, Patrizio. The second question I had is, wondering if you could give us any updates on the vertical power application and how you're progressing towards bringing that technology to high volume production.
We were actually, I was visiting in the Bay Area last week. In a couple of days, I visited eight customers. I think in six out of eight, we were in there visiting the customer because of the customer realization that their future GPU, their future AI ASIC, FPGA, was going to need vertical power delivery in order to meet efficiently current demands ranging up to 1,000 amps and potentially over time, beyond. In some cases, with some of these customers, if we looked at the solution in its entirety as an array of reticles, an array of ASICs that are in proximate relationship to each other, i.e., reticles of a wafer or devices, ASICs, that for latency reasons need to be in very close proximity to one another so that they in effect are above each other. Vertical power delivery is the only way to get there.
We are seeing it, getting traction. We are seeing it as a critical need in order to enable these next generation solutions. Some of them are already approaching, I would say, limited production, not large-scale production, but there's been 1 startup company that made an announcement, I think about a month ago, with respect to very advanced solutions that already relies on vertical power delivery, and for which we are developing a more advanced system with double the current capability. We see it as a common denominator need in that space where, in effect, functionality, computing capacity is clearly dependent on delivering, in some cases, tens of thousands of amperes within a very small footprint, essentially, a foot on the side.
Do you think that's something that goes to higher volume production in the second half of 2020, or is that an application you think is a little bit further out in terms of timing?
It depends on the company and the application. I think there's hope with at least one of these startup companies for volume production in 2020. We're involved in other solutions that involves our so-called GCMs that are due to go into production 2021. There's another application, which is an FPGA type of application that is 2022. You get a range of timelines. Going back to what I learned from the visit in Silicon Valley last week, is that there is a common denominator demand for these kinds of solutions, and we're uniquely equipped and uniquely capable in terms of our intellectual property to support these kinds of solutions.
Thanks, Patrizio. The last question from me, I think it's pretty clear what your advantage is in the high-performance computing and data center markets, where you can get your MCMs closer to the processor than traditional power supplies, and therefore, you have significantly lower distribution losses with your solution. I'm wondering, can you just sort of discuss the advantages you have in the automotive market? Do you have similar advantages to what you're seeing in data center? Are these in the automotive market, more zero current switching, zero voltage switching architectures that some of the other power management suppliers may also have in-house? Just I'm trying to get a better understanding of what's your advantage on the automotive applications.
Without mentioning names, I'll make reference in the back of my mind to one recent design win for which we recently got a first NRE order for development of a solution that involves conversion from an 800-volt bus, which relates to a battery stack, in hybrid vehicles initially, eventually into full EVs, where our solution is a 2-stage solution. It provides, first of all, bus conversion from 800 V. That's a nominal 800 V. It can go higher than that, it can go lower than that, but nominally 800 V, down to 40 V. It's once again the 40-volt intermediate bus infrastructure that I think sets us apart from aspiring competitors. There's a first stage involving a Bus Converter technology, which is way above anything else in the competitive landscape.
Then it's followed by 2 stage, which is our PRM regulators that take that BUS converted voltage level that can track up and down depending on the state of charge of the battery pack, and regulates it to distribute within the vehicle at a regulated 48 volt. We go down from there. There is another part to the system where we take the 48 volt down to 12 volt to provide the power for legacy 12-volt loads that are going to still be there in vehicles for quite some time, because obviously that's a longstanding type of solution that will not be supplanted overnight. They will continue to live for quite some time. Going back to your core question, how do we fit there?
First of all, we fit by converting power from 800 V to 40 V, regulating it at 40 V, and then supplying it selectively to legacy loads that are going to continue to operate for a number of years at 12 V until they go away. This is in the infrastructure of the car. In one application, we're talking about 20 kW, maybe going to even higher power levels. Our solution is extremely dense, extremely efficient, and very cost-effective. That's the value proposition. In another instance, I'll give another example. We're working with another customer that has got autonomous driving needs. It's another modular application. In that case, the value proposition to your earlier point, isn't a conversion from high voltage buses to 40 V. It's instead powering a very current hungry ASIC.
Personally, I don't know exactly, but I presume seven nanometer ASIC, consuming 700 amps, powering it through vertical power delivery from a 40-volt source.
Thank you, Patrizio.
Next question is coming from the line of John. Please go ahead, your line is open.
Hi, Patrizio. I'm wondering if you can give us some more detail on the design wins that you have in-house. Along that line, you mentioned before a 2 large hyper center that we should start seeing bookings for in the fourth quarter, and how these design wins give you confidence that you'll be growing your bookings 30% in the fourth quarter.
We have high confidence that we're going to be able to grow the bookings in the fourth quarter by about 30%. That's a bottoms-up forecast. I'm not going to give you, for obvious reasons, the details of that, but it's coming out of a combination of new design wins and some exceptional demand from older programs. It's a combination of factors at play, but I'm not going to give you the specifics of what makes that forecast come together. We're confident of it, and that's the way I'm going to leave it for the time being.
Can you talk about At one time you had mentioned that we had another hyperdatacenter that might start ordering product in the fourth quarter. Is that still a possibility? Along those lines, are there more hyperdatacenters in the wings? It seems like it's taking a little bit longer than expected for people to switch over to 48 volts, and I'm just wondering if you're seeing an acceleration of that and if you can give us any kind of details on that.
I think the conversion to 48 volt has become evident for everybody to see. Let's put it this way, even our competitors, I think, acknowledge at this point that that is happening and that it is here. It is here in the center space, in accelerators, in supercomputing. As we were discussing a moment ago, it's also coming here with automotive. Ultimately, I think automotive is really the elephant in the room, so to speak, meaning that the scale of usage in automotive systems and the content electrification of automobiles in a variety of ways, drives common denominator standards, not just within that industry, but also in diverse industries, because, again, of common denominator economies. I don't think there is any debate any longer that can reasonably be had with respect to 48 volt becoming a relatively universal standard.
To say it in different words, 12 volt is going away because of limitations that at this point are clearly recognized in all of these end markets. It's not going to go away overnight. As I was suggesting earlier, there's going to be legacy loads for quite some time, but that's not going to be a growth market. The growth market is going to be 48 volt, and at 48 volt, our technology is without peers.
I see that. I guess what I'm trying to get at, though, is it seems like from the outside that you've had one large data center that really has been taking a large chunk of your production. I would expect we're going to see some more shortly, and you had mentioned a couple of quarters ago, or maybe it was last quarter, that there was another one in the wings. I'm just wondering when might we start seeing other larger hyperscalers really coming on board to the extent that you've got the current large hyperscaler guy taking product?
I think it is coming. I'm not going to make very specific references. I'll repeat things I've said in the past. With respect to these kinds of things, I acknowledge I may have been at times ahead of when the future happens, but I think I've been pretty reliable about the nature of the future that is going to come about. I'll persevere in saying that the convergence to 48 is happening. I think developments, for instance, in the GPU space, GPU is going to 48 volt. It is accelerating, no pun intended, that trend. The automotive developments are going to accelerate it further. We are on the verge of having a much broader adoption. It is happening with design wins. There isn't a marquee target customer today that isn't talking to us or asking us for proposals, for solutions for their next generation requirements.
We are there, and our solutions are far superior than anything else out there. In many cases, our solutions are really the only way to enable what the customers need to do. I acknowledge the fact that we've been dependent for quite some time, unfortunately a bit too long, on some very limited number of key customers. Again, as I look at the visits we had in the valley last week, I see all those customers becoming major Vicor customers in years to come. In some cases, it's going to be next year, other cases it's still a couple of years out, but I see it as somewhat inevitable development. As that happens, I think we're going to get the statistical base of business that we've been aspiring to get. As I mentioned in the past, not having it can lead to lumpy performance.
That's unfortunate, but it is part of the innovator role that we've been playing and are continuing to play. I'm very patient with respect to this, very determined. I see it coming. You'll be able to gauge the accuracy of my predictions in the next year or two.
That sounds good. What I think I'm hearing is that you're very confident of the 30% increase in bookings this fourth quarter. There's a very long runway down the road where we can see continued growth in bookings and revenue with all the different design wins and having a statistical base, we shouldn't be as lumpy as we had before. Is that a pretty good summary?
I think I may have used somewhat different words, but I think I would characterize it as a pretty good summary. We see good things happening, starting in the fourth quarter and extending into next year with programs going into production, and more and more design wins that are going through the gestation phase.
Excellent. Thank you very much, Patrizio.
Welcome.
Congrats on the cash flow. That was really nice to see.
Thank you.
The next one is coming from the line of Richard. Please go ahead. Your line is open now.
Well, thanks, Patrizio and Jamie for taking my questions. Maybe a couple of quick ones to start off with here, just to clarify on the bookings that you're talking about for the fourth quarter. Is this driven largely by a single customer, I think in the GPU space, or is there a little bit more balance between some of the other dynamics you referred to as well?
No, there's balance. In fact, the biggest single booking that we anticipate in the fourth quarter is not coming from that space at all. I think on the GPU front, the real ramp in bookings should start in Q1. There may be some in Q4. There is obviously a growing list of design wins that is coming our way. There may be opportunities also in the front end part of the business coming our way next year. There was a recent example that just came up within the last week. It's diverse, to your point. It's not coming from one existing major customer. It's coming from a few different directions.
Okay, perfect. Quick question just on the commentary regarding revenues for this quarter. You said it was difficult to predict. I'm assuming the difficulty is coming largely from the brick business. Is that fair?
Well, I'm not sure that it is just that. I think it's also from advanced products, and our ability to turn within the quarter, depending on where the orders are placed within the quarter, given capacity as the quarter progresses. We're not comfortable making a forecast with respect to providing guidance with respect to revenue this quarter. We're more comfortable giving guidance with respect to the bookings level.
Okay, fair enough. You had some good commentary regarding the automotive space here, both in the kind of traditional powertrain, as well as with autonomous driving. If you look out a few years, Patrizio, what do you think your content opportunity per vehicle could be?
Well, it very much depends on which particular type of vehicle we're talking about.
What if it's a full EV with autonomous driving capability?
By the way, full EV is not necessarily the best scenario. I think hybrids, let's say a hybrid with 20 kW worth of power conversion through a couple of stages, with then some select downstream conversion to 12 V, there's a lot of dollars in that. It doesn't have to be full EV to represent a good case scenario. What I can say is that if it is a complete power distribution infrastructure from a battery pack to a 48-V distribution, including some conversion to legacy loads at 12 V, at the level of 20 kW, I'm not going to say what the number is because it'd be giving away the valuable position in terms of cost per watt, which I'm not going to do. It's safe to say that it's a substantial content. It's very substantial content.
Okay. Well, we look forward to seeing that over time. My last question, or maybe kind of a quick set of questions here on competition. Patrizio, can you kind of characterize above what level of amperes do you think you have no effective competition? Also, have you seen customers who have chosen a solution that is, in your words, not fully enabling? Are you still engaged with those guys if they are?
We've had instances of customers, I can think of one in Japan as an example, that tried to deploy a competitive solution and couldn't make it work. I forget the exact current level of that application. I think it may have been in the 400 amp category. To give you a little bit more technical color with respect to this. The competitive landscape is very consistent. It relies on regulators. These are so-called buck regulators that are capable in 1 stage of delivering, well, they're advertised to be capable of as much as 60 amperes. In reality, because of thermal limitations, they're used at something less than that. If you take one of these devices and in the real world, you can get 30 or 40 amperes out of it.
If you need to support an ASIC that consumes, on a steady state basis, 400, 500 amps, you need a dozen of them. If the ASIC has got transient requirements, peak loads that go 2x, the so-called TDP or steady state current consumption, then with a competitive solution, what you need to do is double up on the number of regulators. Whereas with our solution, we have a freebie, if you will, which is that on instantaneous basis, on a timescale of milliseconds, which is long enough to support these turbo modes of ASICs, the customer can draw double the current without having to add more hardware and pay for it.
When we get to the level of 400, 500 amps with peak loads that could be going up to 800 amps or 1,000 amps, our solution that involves with vertical power delivery, a single Current Multiplier, as you can imagine, is far preferable to a "competitive alternative." They may require 20, 24 phases that don't quite fit, generate a lot of noise. A lot of these advanced ASICs running on very low voltages are noise sensitive. You got the following dilemma with a competitive solution. To limit power distribution losses, the regulators and a large multiplicity of them needs to be deployed close to the ASIC. Well, first of all, there's not enough room to do that. If you could somehow bottom up against the ASIC, you got to worry about something else, which is they generate a lot of noise.
Because of that, the ASIC may not be able to operate nearly as well as it can with our solution, which is fully soft switched with our packaging. There's metal shielding, there is no noise. Fundamentally, it's a very predictable, very high-performance solution that doesn't present both the fit and the noise challenges of the competitive alternative. I would say the threshold, there is probably a band of not a bright line, a demarcation line. I would say at the level of a couple of hundred amps, it is still possible to compete. It gets relatively hard as you progress from 200-400 amps. It becomes mission impossible for the competition above that level. That's where a lot of ASICs are going in many application environments.
Patrizio, that was a great detail there. Thank you for that. That is all my questions.
Thank you. If there is another question, we'll take it. Otherwise, I think we're one hour into it.
Yes, we do have two more questions, actually. We have Doug next. Please go ahead. Your line is open.
Yes. Hi. I won't take a lot of time. I know we're running late. Just two quick questions. I read a report recently that one of your patents that was issued in 2005 on a particular converter, I don't know if it's pronounced sine or scene, Sine Amplitude Converter that expires next year. If that's correct and it expires, is that significant in any way? Then the second part of the question, which follows up to the other guy. When you go and talk to a new customer, potential new customer, what pushback are you getting? Clearly we feel, I feel as an investor, you feel as you communicate to us as investors, that your product is completely superior. Is it price? Are they fearful of production capacity? Do they have doubts about the 48-volt conversion?
Why would somebody try a completely inferior product, like you said, have it fail, only to come back to you again? Those are my two questions.
Let me take the first one. One of our patents, maybe a few of our patents, because there is about 100 patents that are relevant to protecting our IP with advanced products, will be expiring in the next couple of years. You can be sure of the fact that we haven't been sitting still, right? Since 2005, we have continued to make innovations in that particular area. There may be at least half a dozen other patents. The way our products, our Sine Amplitude Converters operate today, and the way they used to operate back in 2005, there's a world of difference. There have been very significant innovations. You can be sure, having been very vigilant with respect to our intellectual property, that we've gone very far in terms of protecting our innovations.
Fundamentally, the way I look at it, the analogy that I like is planting a minefield. Any competitor that wishes to follow our technology and abuse it or usurp it, has got to confront the crossing of a minefield. Maybe one of the mines in the minefield being old gets to be inoperative, but there's plenty of other mines that can blow limbs off, and that's what's going to happen if anybody tries.
I see that. I just wanted to ask.
That's why we're watching this very carefully. We're not seeing, we're seeing something on the fringes, and we're keeping a close eye on that, but it's not mainstream Vicor technology. When it comes to the core technology, I think the industry understands that we're very good, very adept. We invested literally hundreds of millions of dollars in developing this technology, and we're ready to assert the IP. I don't think anybody's got the guts to try. With respect to what happens with customers, nowadays, it's mostly as they come to us looking to implement an advanced solution. I think what goes through their mind, as we discussed in the past, is the appeal of enabling a more advanced system.
A power system that allows their end product to have superior attributes, birthed by the fear, and I would say misplaced fear, that it's a single-source solution, and Vicor is the only game in town with respect to that single-source solution. Now, our track record of supplying a large number of customers and a few key customers abundantly demonstrates that we're a very reliable supplier. We're a very honest supplier. Customers can rely on us for price, for capacity, for delivery, for quality. Not everybody who doesn't know us believes in that. The handicap or the hump they have to get over, has to do with the appeal on the other side of the fence of having a multi-source, quote unquote, commodity solution. That's the dilemma that customers face.
If it were the commodity management team, their choice would logically be, give me a solution that I can buy from a multiplicity of vendors, I can knock the vendors' heads against each other to get lower price, and I can get multiplicity of sourcing with diminished risk, quote unquote. The engineering team and the project managers are attracted to our solution because it's enabling. It gives them the capability to do something that with the commodity multi-source alternative, they can't do. What I'm seeing, what I saw in visits last week in the valley, is that enabling a superior solution wins the day for obvious reasons, because that's more important to the customer. The ability of their end product to compete effectively against their competitors, that's more important than concerns that can be addressed with respect to single sourcing with Vicor.
Customers come to visit our facility. They consistently come away impressed with the state-of-the-art, world-class manufacturing capabilities we have, the control systems that we have in place. Our capabilities reassure them with respect to the fact that. The single source nature of our solutions isn't something they should lose sleep over. Some customers or potential customers do. I think this is the natural tension between, in effect, our product and the commodity competitive alternatives that can't keep up with the level of capabilities we have.
Just real quick, if I understand you correctly, it's sort of you educating them to overcome their fear.
I think that's very important. I can't say that we can succeed. I'm aware of very large customers that have an edict, or at least have had an edict historically, that they're not going to commit their solutions to a single source. Other customers that in the past had similar resolve have crossed the fence. In some instances, their system integrators crosses the fence for them and adopts our solution. It is a consideration. Without it, obviously, the rate of progress would be greater. The tide is turning in our favor. And as we grow larger, I think some of the concerns with respect to size are also going to get dissipated. I think we need to be patient with respect to this.
What we need to do is keep raising the bar and continue to advance our level of performance, leaving the competition further in the dust. That's what we're doing.
Okay, thank you.
The last question is coming from the line of Alan. Please go ahead. Your line is open.
Yeah, good afternoon. Just one question. How are you doing on the new RFM Power Tablet? Is that completed on the new 4G technology?
It's moving forward remarkably. You indirectly referenced it in a comment without being specific about the RFM. Just within last week, there's a customer that has talked to us about very significant volumes for the RFM next year. We'll wait and see. We're also working on completing the next generation of these kinds of products based on our 4G control system that's nicely coming along. We're going to be able to raise the bar, as I suggest in just the last answer to the earlier question, raising the bar with respect to the lower performance of the Power Tablet.
While competitors may be striving to achieve the level of performance of a Power Tablet, which is literally in other money, is smaller than anything else in the competitive landscape, we're far along with respect to a next generation solution that will raise the bar by another significant factor, both in terms of density, efficiency, and most importantly, cost.
Okay. Is the 4G completed yet, or is that still coming?
We are shipping. We actually recently completed qualification of some 4G products. The 4G PFC, which is what's relevant to the RFM or any AC-DC converter.
That's not quite completed yet, but it's approaching completion. We have prototypes on the bench, and we expect it to get to the finish line in another few months.
Okay. The 4G version of the Power Tablet, is that completed yet?
We have prototypes on the bench on a much smaller footprint. The RFM, the Power Tablet, as you might have seen from our website, it's an assembly in a metal shell. Future PFM or RFMs, in other words, AC-DC converters, are going to come right out of our molded panels. We're going to slice them and dice them out of panels, and that is going to make them a lot more cost-effective than the RFM itself is. That's progressing along. We have prototypes on the bench, but the product is not quite ready for prime time. It's still a few months away.
Okay. Thank you very much.
You're welcome. With that, we thank you, and looking forward to talking to you in a few months.
Thank you. Everyone, that concludes your call for today. You may now disconnect. Thank you so much for joining, and have a great day. Goodbye.