Good afternoon, and welcome to Vicor Corporation's earnings call for the third quarter ended September 30th, 2021. I'm Jim Schmidt, Chief Financial Officer, and I'm in Andover with Patrizio Vinciarelli, Chief Executive Officer, and Phil Davies, Vice President of Global Sales and Marketing. After the markets closed today, we issued a press release summarizing our financial results for the three months ending September 30th. This press release has been posted on the investor relations page of our website, www.vicorpower.com. We also filed a Form 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 make during this call may constitute forward-looking statements for the 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, and our capacity expansion, as well as management's expectations for sales growth, spending, and profitability, are forward-looking statements involving risk 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 in or implied by any of our remarks today. The risks and uncertainties we face are discussed in Item 1A of our 2020 Form 10-K, which we filed with the SEC on March 1st, 2021. This document is available via the EDGAR system on the SEC's website.
Please note the information provided during this conference call is accurate only as of today, Thursday, October 21, 2021. 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 today's call will be available beginning at midnight tonight through November 5th, 2021. The replay dial-in number is 888-286-8010, followed by the passcode 33342563. This dial-in and passcode also are 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 now turn to a review of our Q3 financial performance, after which Phil will review recent market developments, and Patrizio, Phil, and I will take your questions.
In my remarks, I will focus mostly on the sequential quarterly change for P&L and balance sheet items and refer you to our press release or our upcoming Form 10-Q for a year-over-year comparisons. As stated in today's press release, Vicor recorded total revenue for the second quarter of $84.9 million, down 11% from the second quarter total of $95.4 million. Revenues came in substantially below expectations because of semiconductor component shortages compounded by our own capacity constraints. Component shortages were due to limited wafer allocation and long cycle time back-end semiconductor component packaging processes. Advanced products revenue rose 6% sequentially, while brick products revenue declined 23.7% from the second quarter. Shipments to stocking distributors declined 36.6% sequentially, primarily due to a decline in brick products.
Exports for the second quarter decreased sequentially as a percentage of total revenue to approximately 62.4% of consolidated revenue from the prior quarter's 64.3%, primarily due to decreases in brick products. For Q3, advanced products share of total revenue increased to 51.2%, compared to 43% for the second quarter, with brick products share correspondingly decreasing to 48.8% of total revenue. Turning to Q3 gross margin, we recorded a consolidated gross profit of 50.4%. Gross margin dollars declined by 14% sequentially due to the decline in volume, but increased 28% from the same period a year ago. Margins remain under the pressure of high tariff charges, though the Q3 charge was approximately the same as Q2's charge of approximately $1.9 million. We expect to see improvement over time, in part reflecting our ongoing efforts to reduce component imports from China. I'll now turn to Q3 operating expenses.
Total OpEx increased 3.3% from the second quarter, driven by increased compensation, legal and consulting fees, and recruitment expense. The amounts of total equity-based compensation expense for Q3 included in cost of goods sold, SG&A, and R&D were approximately $259,000, $1,033,000, and $575,000, respectively, totaling approximately $1.9 million. For Q3, we recorded operating income of $12 million, representing an operating margin of 14.1%. Turning to income taxes, we recorded a net tax benefit for Q3 of $886,000, representing an effective tax rate for the quarter of -7%. The net tax benefit for Q3 and year to date was primarily due to a result of the income tax accounting required for stock options exercised during this period. Net income for Q3 totaled $13.3 million. GAAP diluted earnings per share was $0.29 based on a fully diluted share count of 45,034,000 shares.
Before I turn to our financial position, just a brief update about COVID-19 and our workforce. As previously discussed, as a designated essential manufacturer, using masks and practicing social distancing from the onset of the pandemic, we have continuously operated three shifts at our Andover manufacturing facility. Cases and absenteeism due to COVID-19 are now negligible. Nevertheless, because much of the potential influence of the COVID-19 pandemic is associated with risk outside of our control, we cannot estimate the extent of such influence on our financial or operational performance, or when such influence might occur. Turning to our cash flow and balance sheet. Cash, cash equivalents, and short-term investments totaled $228.9 million at the end of Q3. Accounts receivable net of reserves totaled $51.1 million at quarter end, with DSOs for trade receivables basically steady at 40 days. All balances are current.
Inventories net reserves increased 11% sequentially to $63.4 million. Annualized turns decreased slightly to 2.91 from 3.12 for Q2. Operating cash flow totaled $10.1 million for the quarter. Capital expenditures for Q3 totaled $15.2 million. We ended the quarter with a construction-in-progress balance of $36 million, leaving approximately $20 million of our capital budget scheduled to be spent through the year. Our factory expansion is proceeding on schedule and on budget. I'll now address bookings and backlog. Q3 book-to-bill came in at 2.0, and one-year backlog more than doubled from the same period last year. Turning to our outlook for the fourth quarter of 2021. Our practice continues to be to not provide specific quarterly targets. However, given our assessment of available semiconductor components and of our capacity, we are planning on a 20% increase in revenue in Q4.
Availability of semiconductor components has improved since Q3 because of reduced cycle times in back-end processes and a substantial increase in our wafer allocation. Easing supply chain bottlenecks and recent increases in our advanced products manufacturing capacity should enable a significant step-up in advanced products revenue. We continue to focus on improvement in product-level profitability. Further, we do not anticipate any meaningful increases in operating expenses. While substantial further improvements in gross margin will have to await production from our new vertically integrated factory, we expect incremental revenue to drive earnings per share, given the scalability of our operating model. With that, Phil will provide an overview of recent market developments, and then Patrizio, Phil, and I will take your questions.
I ask that you limit yourselves to one question and a related follow-up so that we can respond to as many of you as we can in the limited time available. If you have more than one topic to address, please get back in the queue. Phil?
Thank you, Jim. Good afternoon, everyone, and thank you for joining us. In Q3, we achieved record bookings, a record book-to-bill, and record backlog. Bookings for AI and data center customers for Factorized Power Solutions and 12V - 48V bridging bus converters were robust, and our visibility into production programs is longer than we have seen on previous projects. Some of the next-generation programs will also run in parallel with existing programs, which is a new strategy for some of our data center customers. Next-generation AI processor and data center CPU server programs in early stages of development and pre-production will provide $70-$150 of Vicor content per board. New clustered processor AI applications offer much higher dollar content for our proprietary Vertical Power Delivery solutions, whose IP protection and technical challenges set us further apart from our competitors.
Expanding our SAM within the large data center market is always an objective for our teams. An evaluation of our advanced AC-to-DC solutions have started at lead customers with derivative products in development that will target applications for both single-phase and three-phase AC-to-DC across the high-performance computing, industrial, and aerospace and defense markets. We are also engaging with customers who are deploying ASICs for high-speed optical networking units, which have power delivery challenges due to major increases in core rail currents above 1,000 amps. Our first major customer in this new market will be in production in Q3 2022. As the data center industry shifts to 48 V based rack power delivery, the upcoming Open Compute Project will feature 48 V power distribution and the need for a more robust ecosystem of high-density modular power solutions rather than low-density discrete approaches.
All of which validates decisions we made to commit to this market before it emerged 10 years ago. Our OEM licensing practice will facilitate an ecosystem of high-performance power system solutions for the AI and data center industry. Anticipating power delivery challenges and trends and innovating to deliver solutions ahead of market needs is Vicor's track record. This is playing out with additional growth in Q3 for our pipeline of automotive opportunities as electrification commitments grow within OEMs for additional models and their new electrified vehicle introduction dates solidify. We are not only seeing more opportunities in pure EV, but also in plug-in and mild hybrid platforms, as well as the truck industry, broadening and diversifying our market and customer base. With this backdrop of increasing growth opportunities, we have decided to restructure the front end of our business into market-based business units.
The four business units will be high-performance computing, automotive, aerospace and defense, and broad markets. This structure aligns with our Five Layers of Growth strategy and brings a higher level of focus on a global scale to our target markets, customers, and applications, as well as increased responsibility in the team for funnel conversion, revenue streams, and gross margin improvement. Because of the high level of technology reuse and applicability of our innovations across markets, engineering and product development will remain centralized. Patrizio, Jim, and I will now take your questions. Thank you.
Okay, operator, we're ready for you to coordinate questions for us.
All right. Absolutely. Everyone, if you wish to ask a question, please press star one on your telephone. We have two upcoming questions already. The first one is coming from Jonathan Tanwanteng. Please proceed.
Hi. Good afternoon, guys. Thank you for taking my questions. That's a really nice number you have on a sequential increase in Q4. I was just wondering, how much of that is just catch-up from Q3, and how much of that, I guess, is organic demand? I guess that is to the upside of maybe what you had been thinking when you had said 7% sequential increases over the next several quarters way back in Q1 and Q2.
It's all about demand. Our backlog as we entered the fourth quarter without any allowance for turns business would have us up by approximately 20%, and you would expect that turns business would layer on top of that a substantial incremental amount. The demand is there. We're still capacity constrained, and there are still challenges with respect to, in particular, semiconductor components. As Jim articulated in his prepared remarks, those challenges have to do with both the back-end processing after completion of wafer fab, and they have to do with wafer allocation. What has improved considerably since Q3 is that the back-end bottlenecks, which were in part related in some parts of the world to COVID-19 and other factors, those have been largely relieved.
The components that did not show up in time because of back-end bottlenecks in Q3 have shown up or are about to show up in time for our builds in Q4. Furthermore, we have recently been able to receive commitments with some of our foundries that result in considerably greater allocation, both within Q4 and more significantly into 2022. We believe that in terms of the component gates that in effect either launch within the third quarter in terms of compounding capacity constraints, as we all know, were already there. We made a great deal of progress. The capacity constraints are still there, but being able to start builds on scheduled dates and not having to be as much hand-to-mouth with respect to the availability of components will facilitate a substantial capacity increase this quarter.
Got it. Thank you, Patrizio, for that color. Second, I saw in the press release that it was mentioned that you signed a license with some of your customers, or at least one of your customers who was buying these infringing product. Can you talk a little bit more about that? Does that set the tone for the other customers out there that are doing this, and what kind of benefits do you expect to see from this?
We're not at liberty to say anything regarding the identity of the licensee or any of the key parameters of the licensing deal that we executed in the third quarter. I think in general terms, I can tell you that it points to the beginning of new kinds of relationships on the licensing front. The industry, as we all know, has growth needs and technology needs that are very compelling.
Needless to say, particularly at the time of capacity constraints, there is a legitimate interest on the part of customers to be able to secure their requirements. To the extent that enabling technology has only been available from Vicor, that's something that customers would like to, in various ways, be able to overcome. The OEM license provides customers with the ability to, without concerns with respect to possible exclusion orders to the importation of OEM products into the U.S., procure modules, incorporate those modules into systems, even though those modules would otherwise infringe Vicor patents. We believe there's going to be more licenses, and we view this as a complementary component of our business model to contribute to margins profitability, while providing the industry with the ecosystem that many key customers seek to have.
Okay, thank you. I'll jump back in queue.
Thanks, John.
The next question is coming from Quinn Bolton. Please go ahead.
Hey, guys. Congratulations on the nice bookings activity. Before I get to my bookings question, just wanted to come back to your outlook for the fourth quarter. Obviously, component availability and your manufacturing capacity are significant constraints, and I'm just curious, do you guys feel confident that where you sit today, you have enough component availability and internal manufacturing capacity to grow the business 20% quarter-on-quarter?
We do. I want to be clear that this is not a slam dunk, to put it figuratively, in that we're at the beginning of the quarter and a lot has to happen. It's not that we have as of today, all the semiconductor components we need in order to make the revenue plan for the quarter. They are on their way, and they'll be on their way through the first two-thirds of the quarter. There are still risks out there. We feel that the improvements that have been made with respect to the back end, which in effect made available components that were destined to us in the third quarter but couldn't be processed in time.
That coupled with, as I mentioned earlier, the ability to achieve greater utilization of constrained capacity by not being gated by the inability to start because of lack of particular components. That sets us off to a much faster start early in the quarter, enables well with respect to the outcome of the quarter as a whole.
Thank you, Patrizio. My second question is just on orders. I think most of us on the line that have followed the company have been very impressed with the acceleration and the bookings number, and I think most of that's being driven by advanced product bookings, where I think advanced product bookings last quarter was probably over $90 million. This quarter, it looks like you booked over $100 million, maybe as much as $120 million in advanced product bookings. My question is, how broad-based is that advanced products booking activity over the past couple of quarters? Is it highly concentrated among just a couple of GPU, TPU, ASIC accelerator customers, or are you starting to see very good diversification, including front-end or AC-DC, including the NBM modules? Can you just give us some color on how broad-based that bookings activity is? Thank you.
Hi, Quinn. This is Phil. As we've been talking over, I think the last 18 months, the customer base for advanced products in data center AI has been broadening quite a bit from one or two major companies a couple of years ago to now well over 12, 15 type quantity of customers, which are adding to the backlog and the bookings number. In terms of the activity there, 48- 12, as well as the bridging 12 - 48 and 48 - 12 on some of the CPU server boards, it's very strong. We're entering the phase of continued AI growth with our existing customer base, but also with the new Intel and AMD CPU platforms coming on in data centers with some of the hyperscalers who have gone 48 Vs. We're seeing strength there.
Also coupled with the AI accelerator growth, which are 48 V based, we've seen growth now, significant growth in our 12- 48 V bridging converter business, because hyperscalers that are still at 12 Vs in their infrastructure need the 12 - 48 to use the latest, greatest accelerator cards and multiple cluster AI accelerator cards. It's very broad spread.
A quick follow-up. Is it pretty well balanced between the types of MCM or MCD components versus the NBM bridging products, or does it skew more heavily to the point of load MCM, MCD components today still?
I think the 48 to load is the largest piece of it. The NBMs are doing really well in the 12- 48 space. One of the reasons being is that typically in those applications, you've got six to eight NBMs per power board. The quantities add up pretty quickly, and the bookings add up pretty quickly.
Got it. Thank you for the color.
Great.
Just a bit more on that. The NBM market opportunity next year is quite substantial. Looking at it from the timing perspective, it hasn't been all that significant, but it promises to be. In fact, it has been with recent orders, just within the last couple of days, a $4 million order for NBMs, quite significant as we enter in the first half of next year and the middle of next year.
Thank you.
Thanks, Quinn.
The next question is coming from Richard Shannon. Please go ahead.
Well, hi, guys. Thanks for taking my questions. I might follow up on the topic of licensing. Patrizio, while I understand you're not allowed to say much about this, maybe I'll ask a couple of questions and see if you can push the envelope here. Can you talk about any particular applications that the said licensee is working in? Then also, in terms of the nature of the agreement, whether there's upfront and ongoing royalties or just upfront, just any nature of understanding there would be great, please.
Again, I need to keep it at a very general level, that would not in any way infringe on the confidentiality constraints we have with the licensee and future licensees. The license structure is fundamentally a pay-as-you-go, and the mechanism involves the placement of the license POs based on the licensee's choice to get a license for particular products to be used in particular systems. It's incumbent on the licensee to decide for which product, exactly product that may otherwise infringe Vicor patents, to get a license. As mentioned in the press release, not only did we get a first OEM license executed within the quarter, but from that licensee, we have received license POs and in fact, revenue, even though those POs will result in cash flow starting this quarter. That's, I think, a representative example of licensing opportunities to come.
We have a well-thought-out methodology for this. Potential licensees can take their time to decide whether or not they want to secure a license. They can have a wait-and-see attitude with respect to that, but there are serious risks of a potential exclusion order that could affect their systems, which are quite valuable. That should motivate them to seek a license earlier than later, also because royalty terms escalate with time. It's accepted there is a passage of time that brings about a transition from what we call phase 0 to phase 1 to phase 2. With that comes an escalation in the royalty rate. We believe that having accomplished the first licensee, there's going to be more.
We're aware of a number of OEMs that we believe will need a license if they want to continue to access the kinds of products they need for their systems.
Okay. Patrizio, thanks for all that detail. My second question is on gross margins, not only for the fourth quarter, but I guess looking at it in the next few quarters here as well. Jim, I think I heard you correctly in your prepared remarks. You seem to be suggesting a flattening of gross margins here in the fourth quarter, and until we see some throughput in sales from products in the new facility, it wouldn't go up until then. I want to make sure I'm understanding the dynamics here that you were suggesting in your comments and how we should look at gross margins in the next few quarters.
Sure. First of all, I do have to say that it's our practice to not provide any specific guidance on gross margin in out quarters. I didn't mean to imply by that statement about waiting until the new factory's online that we have to wait for gross margin improvement per se. Volume increases drive gross margin leverage. That's a fact in the industry. That's actually not just true to Vicor, it's true for anybody. That comment is actually a comment that I made last quarter as well. It's just a standard comment we want to make that highlights the fact that we're all very anxious to get our own factory online. As that comes online, we get all kinds of efficiencies. There's no outsourcing of some of the process steps. We get more leverage as the volume goes up, greater efficiency inside the factory.
That's really going to be a significant milestone for us.
I'll add to that looking at it in terms of the past, the rearview mirror, I'm not saying anything that wouldn't be expected, but it bears saying, in Q3, the gross margins and the profit margins would have been substantially higher were it not for the fact that we had the shortfall in revenue that cost us several points of margin within Q3, right? That's an unfortunate fact. To Jim's point, rising revenues immediately will bring about better margins even before we get the full leverage of the new facility.
Okay, great. Thank you for all the detail, guys. That's all for me. Thanks.
Thanks, Richard.
The next question is coming from John Dillon, please go ahead.
Hey, congratulations, guys, on the bookings and the backlog. Really great numbers. Phil, I do have a question on the bookings number. I'm just wondering if the increase in bookings was due to customers securing a place in line or customers priming the pump, or are those bookings numbers sustainable?
The programs that we're involved in now, John, and the visibility we have, we're seeing increased demand for the AI accelerator cards across the data centers. The deployment of those cards is significantly increasing, with all the hyperscalers on a global level. That's a very strong, robust business. As I mentioned, along with that increase, a lot of the rack systems are still 12V based, so they need the 12V - 48V to power those cards. We're getting sort of a double whammy, if you like, by having the best high-density bus converter in the industry available today. We're winning on both ends, as it were, with both the hyperscalers deploying the AI and also with the guys actually with the AI cards. It's strong right now.
I think what I'm hearing is the bookings are sustainable and even growing.
Yep.
You see that in the future?
Advanced products, yeah, absolutely. Yeah.
Along with that, I would imagine you have pretty good pricing power. Are you guys raising prices?
Well, we're always looking to price on value, right? That's what we do. We look at the value that we bring to our end customers, and we do that with the whole portfolio. It isn't just on advanced products, but even on our brick or legacy business, we adjust prices regularly and, again, we do that on the case-by-case basis of product family or market or customer, but we really do try to price on value.
Gotcha. The follow-up would be, you mentioned the 12 V-48V , and I'm sitting there scratching my head. Why would customers want to do that? Why are they going from high volt age down to 12 volt age back up to 48? Patrizio, you kind of mentioned this in the last call, but aren't you going to see a huge conversion to all 48V racks, or do you see that coming?
Well, that is coming, but it doesn't happen overnight, right? Fundamentally, key developments, turning points in the industry, bring about adoption, but that adoption requires time to get implemented. Needless to say, the absurdity of, to your point, going all the way down to 12V to go back up to 48V, to go down to 1.8 V or sub-1 V, as the case may be, that will provide even greater motivation for the balance of the hyperscalers to convert to 48V racks. There's plenty of motivation, but not just coming from the absurdity of unnecessary stages of conversion to volt ages that are too low for efficient power distribution back up to volt ages that are necessary in order to achieve what needs to be done downstream within the rack.
Yeah, John, this is Phil.
Data centers are going to go through major changes over the next three to seven years. If you just look at the amount of AI that's being added, they're just not equipped to get the data in and out at the speeds that they want, so there's huge networking changes that are going to come in terms of moving from GB to TB of moving data in and out to support autonomous driving and other machine learning applications. The data center footprints that have all of the renewables to power them, because you're talking tens of MW for these things. You can't move a data center to a different location. You've got to refurbish it. You've got to take it to the next level of capacity and capability. All of this stuff is going to change, including 48 V in the rack, more AI, higher performance optical networking.
It's really going to go through a significant change over the next three to seven years. It's going to be quite amazing to watch.
Another element of the engineering of the racks, beyond the electrical part of it that has to do with switching over from 12V distribution to 48V distribution, has to do with thermal management. The power levels within the rack that justify or motivate a transition to the most efficient power distribution volt age, the 48V, which is in reality 54 V, they still safe. That motivation, it brings about also thermal management challenges that before too long, will have a number of these racks be cooled by liquid cooling as opposed to air. In that regard, we should point out that within the third quarter, recently, we have delivered first units of a very high density liquid-cooled front end to a lead customer for that particular capability.
This solution, again, liquid-cooled, three- phase in, 52V, 54 Vs out, is an one order of magnitude more dense than anything else out there.
Guys, this is really exciting. Thank you so much. I'll get back in the queue. Great quarter. Great visibility. Thank you.
Thanks, John.
The next question is coming from Gus Richard. Please go ahead.
Yes, thanks for taking my question. I just wanted to circle back on your constraints. Is the plating capacity still a constraint? Do you need the new facility to come up to relieve that?
To your point, that's the key capacity bottleneck. We are complementing the capability of our plating partner by, to a high degree deploying our own human resources within their facility, to enhance the available capacity. We have also recently taken steps to bring about another partner, actually closer to Vicor main manufacturing facilities. In a matter of two weeks, we expect to turn that second partner on for further step up in capacity. These are stepping stones to the deployment of very high volume state-of-the-art plating capability within our own facility. By the way, when one thinks of plating, it's tempting to think of something that is relatively low tech. That's far from being the case here, right.
What we're referring to is about $25 million worth of equipment on 40,000 sq ft, which will be state-of-the-art for this class of products.
Got it. That is very helpful. Thank you. On the gross margin pressure, it sounds like it was primarily absorption, but I was wondering if there was any influence from higher input costs and/or mix.
There's no real pressure from mix. It turns out that our brick business is actually fairly decent gross margin. Higher input costs, what we've done there, generally speaking, even before I arrived at Vicor, we were not accepting inflation costs in our raw materials without passing those costs on in many cases to our target. We're not absorbing that. I hope that answers your question.
It does. Thank you. The last one from me is just looking at the backlog, and given the supply constraints in the world, are you seeing any aging of that backlog? Is some of it's scheduled further out than normal, or has the distribution backlog remained the same over the last few quarters?
We have obviously implemented longer lead times, so we have a lot more visibility. It ranges from 20-30 weeks for different types of product families. The backlog is, I get asked, people double ordering and all that, and we don't see that. It's real demand. We have a very unique product. It's not a commodity product. There's not a pin-for-pin compatible type of a market for our stuff. We have very dedicated, unique customers that buy our stuff. They have done for many years on the legacy side. On the advanced product side, we're building a whole new business there in data center and AI, so we have good visibility into that and the end market demand and the growth there.
I'm very confident in the backlog that we have in terms of the validity of it and the reality of it.
To be clear, it's within 12 months, right?
Yes.
Everything is scheduled within 12 months.
Yes.
That's what we define to be the backlog.
Got it. Perfect. Thank you so much. That's it for me.
Thanks, Gus.
We have another question from Jonathan Tanwanteng. Please proceed.
Hey, guys. Thanks for the follow-up. Just another question on the capacity and the constraints that you have. I get that you're putting a lot of good plating capacity into place this quarter, and especially the new facility next quarter, in Q1. I was just wondering how confident you are in increasing your wafer allocations and the other components in the chain that you need as we get into the first half of next year, because I know in many other places, these things are getting worse and not better. I was just wondering what your visibility into the things that you can control are as we head into next year.
We have visibility, and we've been allocated a significant step-up in wafer outs in November, and even more so in December. The numbers for next year are not cast in concrete yet, but we have been informed that we're going to get the numbers that we need. As you can imagine, we do have quite a bit of leverage with respect to getting wafer capacity because of the critical programs that our chips support. That's obviously a factor at play with respect to getting access to what's necessary for us to be able to make the products that make AI and data center systems operate.
Got it. That's helpful. I was wondering if you could give us an update just on the auto business. How many platforms do you have at this point? Are there more coming into the pipeline? Just give us an update on, I guess, when you think that'll become significant for you as a source of profit and revenue.
Again, we're starting production at the end of 2023 or mid-2023 with some early customers there. We have more collaborations. That's what I was talking about in my remarks. We've got more collaborations that occurred in Q3. Those collaborations typically take six to nine months to work through in terms of building prototype systems and testing them, and then moving the customer really to a stage where we call the conversion stage, right. Which is where we move to a commitment for a start-of-production date. That's what we're always shooting for. At this point in time, we have two large OEMs with start-of-production dates, and we have a whole host of collaborations that are moving towards SOP dates in the next six to nine months. It's a very exciting funnel we have in automotive.
Got it. Thanks. Again, congrats on the big backlog.
Thanks, John.
We have a question from Alan Hicks. Please proceed.
Good afternoon. You've invested over half a billion dollars on these new products. It now looks like you're going to be solidly profitable for as far as the eye can see. Where are you at in terms of tax loss carryforwards? When do you expect to get back to a normal tax rate?
Alan, Jim Schmidt here. At this point, we're not going to comment on next year. I'll tell you that the current year is status quo, so we expect the tax rate, I know it's toggling between 4% or -7%. We have a tax benefit oftentimes because of stock option exercising that's going on. We're not going to comment at this time on next year, but stay tuned. Maybe next call, I think we'll be more in a position to talk about that going forward. That's what we prepared to say now.
Okay. Where are you at in terms of your tax loss carryforwards? I think it was in the $30 million range.
I think that's right, yeah. That's still on the books. It's a fairly complicated topic. We're working with our audit partners on that, and we're ready to say status quo for the balance of the year. After that, we'll talk again in January.
Okay. I would anticipate that you'd probably use that up next year, I would think, if you're going to make at least $30 million in profits next year.
It's more complicated than that.
Yeah.
I have people on my staff running the numbers routinely, and there's stock option exercises that are a tax benefit and other pieces of the puzzle in there. Yeah.
Okay. Also there was talk a few quarters ago about getting some of the tariffs that were paid, clawing them back. I think they were, if I remember, six months ago, they were $4 million range, I think, and been growing since.
Well, what we said in this.
That's still possible?
Mike, for clear, we said the tariff expense in the P&L and cost of sales was $1.9 million, which was approximately what it was last quarter. We are making progress on reducing tariff expense. It's actually significant amount of time people are spending on that activity right now, and we did make progress. It's incremental. Sometimes it's a bit slow in coming because we're talking about product from Uncle Sam, but we're making progress on it.
Okay. You still expect to get that money back, or a good portion of it?
Yeah.
It's not a matter of if, it's a matter of when.
It's a matter of when. Yes.
Yeah. Okay. Good. Okay, thank you very much.
Thank you.
We have another question from John Dillon. Please proceed.
Hey, guys. Patrizio, you said that you've got some front-end products I think you delivered this quarter, but I'm wondering if we could get some more color on all the front-end products. Are you guys actually actively booking orders for that? How is it going? Are they going into the same markets? Phil, maybe you can give us a little color on that. How are you doing?
As of now, we have 2 major programs that share common denominator, 3-phase input, high power, 48V, 52 V, 54 V outputs. These programs are representative of general market needs that we are gearing up to address. The lead customers provided us with an opportunity to do what was necessary to, as suggested earlier, deliver state-of-the-art solutions. In one case, it's natural convection cool. It's a solution that doesn't require either fan or liquid cooling, which is in effect at one extreme. In the other case, as I was suggesting earlier, it's an extremely high density solution, which is liquid cooled and plugs into a rack system, delivering 20 kW in a footprint that is about the size of an iPad. Just imagine a very thick and very heavy iPad, but in the X and Y footprint, 20 kW going from three-phase AC to 52V, 54 V.
That's the kind of solutions that our customers are going to need in years to come in order to get their racks from the low tens of kW to the much higher power levels that as suggested by Phil earlier. They're going to need in order to get much greater bandwidth within the footprint of their existing facilities.
Yeah, John, this is Phil. We haven't really stepped on the gas yet with that product. We've obviously got some early customers. We're working on some derivatives. As I mentioned, single phase and three-phase, but I think it'll be Q2 probably next year before we really put our foot on the gas and start moving with that stuff.
Q2, you'll start the bookings aspect, that's what it sounds like.
Yeah. You'll see a lot of marketing from us, a lot of activity in the customer base worldwide. We're really going to put our foot on the gas with this stuff because it's very special.
Yeah, it's very exciting. Along those lines, Phil, you talked about the bookings, the numbers, and you see them increasing. If I do my math right, you guys are going to be filling up that new factory very quickly. You must be looking at a new factory. I'm just wondering, what are your plans for a new factory? What's the timing of that new factory?
Jim and I were together with Mike, our head of operations, visiting one such candidate just within the last 10 days. Frankly, I feel that it is still premature to take that step. I believe that we have quite a bit of capacity within the expanded factory. I've had two tours of the expanded factory within the last 10 days. I've been very impressed with what the team has done in terms of laying the foundation for the expansion. By the way, it wasn't mentioned earlier, but it should be mentioned. Within the third quarter, a lot of equipment got relocated within the existing walls of the factory. In fact, within the first week of the third quarter, we had no throughput because we actually took the whole factory down to make new facilities for electric power delivery.
There's a lot of work that's been going on within that factory to set the stage for a level of capacity that I think may well exceed the goals that the team has set and we've reported. With that, I think we got a little bit of time to take the right step for the next incremental capacity. I think it's critically important that over the next six months, our focus be strictly on realizing the major capacity expansion is about to come online. Starting with a contribution to capacity in the first quarter, and then a bigger contribution to capacity in Q2. I think once that has happened, then we can take the next step.
Great. It sounds like you're going to beat your goal of $750 million with the new wing. It looks like you can beat that number.
I think we can. I think that this factory will be very efficient, and it's got more capacity potential that we have targeted. Some of that has to do with the mix of the products. A case in point, our budget in terms of capacity was predicated on some of the early chips that are relatively thick. They take more time to process than thinner chips. A lot of the design wins going forward are going to be for thinner chips. Those thinner panels, you can think of them in some respects as wafers with fewer layers that can process faster. I think we have a variety of degrees of freedom at play to improve efficiency, improve capacity, and being focused on that at this point in time is more important than securing the next increment of brick-and-mortar.
Very good. Love it. Thank you so much. I'm looking forward to it. It's really exciting times. Great job, guys.
Thanks, John. I think we're going to have to close out now. Operator, we're going to have to close the call. Thanks everyone for participating.
Thank you very much, everyone. That concludes your conference call for today. You may now disconnect. Thank you for joining.