Aehr Test Systems (AEHR) 2022 Q4 法說會逐字稿

Good day, and welcome to the Aehr Test Systems Fiscal 2022 Fourth Quarter and Full Year Financial Results Conference Call. (Operator Instructions) Please note this event is being recorded.

I would now like to turn the conference over to Jim Byers of MKR Investor Relations. Please go ahead.

Thank you, operator. Good afternoon, and welcome to Aehr Test Systems Fiscal 2022 Fourth Quarter and Full Year Financial Results Conference Call. With me on today's call are Aehr Test Systems' President and Chief Executive Officer, Gayn Erickson; and Chief Financial Officer, Ken Spink. Before I turn the call over to you, Gayn and Ken, I'd like to cover a few quick items. This afternoon, right after market closed, Aehr Test issued a press release announcing its fiscal 2022 fourth quarter and full year results. That release is available on the company's website at aehr.com. This call is being broadcast live over the Internet for all interested parties, and the webcast will be archived on the Investor Relations page of the company's website.

And I'd like to remind everyone that on today's call, management will be forward-looking statements today that are based on current information and estimates and are subject to a number of risks and uncertainties that could cause actual results to differ materially from those in the forward-looking statements. These factors that may cause results to differ from the forward-looking statements are discussed in the company's most recent periodic and current reports filed with the SEC. These forward-looking statements, including guidance provided during today's call, are only valid as of this date, and Aehr Test Systems undertakes no obligation to update the forward-looking statements.

And with that said, I'd like to turn the call over to Gayn Erickson, President and Chief Executive Officer.

Thanks, Jim. Good afternoon, everyone, and welcome to our fiscal 2022 fourth quarter and full year's earnings conference call. Thank you for joining us today. Let's start with a quick summary of the highlights of the quarter and fiscal year and the momentum we're experiencing in the semiconductor wafer level test and burn-in market, and then Ken will go over the financials in detail. After that, we'll open up the lines to take your questions.

We're pleased to report very strong growth for fiscal 2022 with record revenue for both the fourth quarter and full year. Total revenue for fiscal 2022 was $50.8 million, our highest annual revenue on record and more than 3x last year's revenue. We also had record bookings for the year at $60.2 million. Importantly, we're seeing the significant leverage in our operating model as evidenced by the strong profit for the fiscal year. We also improved our balance sheet significantly with our year ending cash position of $31.5 million and no debt.

Let me go ahead and start and talk about the wafer level test and burn-in market, silicon carbide for electric vehicles, that's where a lot of the excitement is going on. Our strong revenue growth in fiscal 2022 was driven by the demand for our wafer level test and burn-in solutions, particularly for wafer level stress and stabilization of silicon carbide devices for use directly in electric vehicle market. The silicon carbide market for electric vehicles and its supporting infrastructure requirements are growing at a tremendous rate, with Canaccord Genuity estimating that wafer capacity will increase from 150,000 6-inch wafers in 2022 -- 2021 to over 4 million 6-inch equivalent wafers in 2030 just to meet the electric vehicle market alone. This represents growth of over 25x the wafer starts just for electric vehicles. They also forecast another 4 million 6-inch equivalent wafers to address other markets, such as industrial and solar power conversion.

Our lead customer for silicon carbide wafer level burn-in made significant investments in their silicon carbide production throughout this past fiscal year. Today, we're excited to announce that we received $12.8 million in new orders from them for multiple FOX-XP systems, a high-volume production WaferPak Aligner, and a small number of WaferPaks full wafer contactors to meet their increased production capacity needs for silicon carbide-based power semiconductors for the electric vehicle market. All of this is expected to ship by the end of our fiscal third quarter ending February of 2023. This adds to the backlog of systems that we're shipping to them this fiscal year or fiscal quarter, actually. In addition to the system capacity order, we expect significant subsequent orders for WaferPaks needed for the system orders announced today, and they will ship at approximately the same time as the systems.

We have continued to optimize the shipment and installation processes of our FOX-XP systems with the last system shipped to them installed and released fully into production within 8 days after we shipped it out of our factory, and that includes international shipping. We're excited to see them continue their ramp, and we continue to expect significant additional system and WaferPak purchases from them over the next several years and through the end of the decade as they strive to be a market leader -- market-leading supplier of silicon carbide devices.

Aehr Test provides a highly unique and cost-effective solution for applying the stress test across every device on an entire wafer before they are singulated and put into packages or multi-chip modules. This allows our customers to burn-in every single device at a lower cost than they could in any other form due to our ability to contact thousands of devices on a single wafer and test 18 wafers in a single system with our FOX-XP multi-wafer testing burn-in system and proprietary FOX full wafer contact WaferPaks.

Silicon carbide-based devices such as MOSFETs, which stands for metal-oxide-semiconductor field-effect transistors, are extremely efficient, rugged and reliable semiconductors that are used in the power conversion to charge the batteries and battery electric vehicles and in the engine controller traction inverters that drive the electric engines in EVs. Industry-leading semiconductor suppliers like our lead silicon carbide customer tell key differentiators of silicon carbide over the silicon-based IGBTs, which are insulated gate bipolar transistors that include silicon carbide's higher system-level efficiency, owing to the greater power density, lower power loss, higher operating frequency and increased temperature operation.

This translates into a higher driving range on a single charge, smaller battery sizes for traction inverters and faster charging time for onboard chargers. These silicon carbide-based traction inverters were first used by Tesla in their Model 3 electric vehicle sedan. Tesla then converted all their electric vehicle engine controller traction inverters from silicon-based to silicon carbide-based MOSFETs. Forecasters like Canaccord Genuity, yield development and actually what, believe that most traction inverters will be silicon carbide base within the next several years.

One of the biggest concerns of existing and would be electric car buyers is range anxiety, the fear of running out of energy before making it to a charging station. As some of you may have seen, just recently in all electric Mercedes-Benz, recently drove 747 miles without recharging, handily beating every electric vehicle on the market today. This ultra-long-range electric car called the Vision EQXX was sponsored by ON Semiconductor, onsemi and highlights the significant progress that is being made with these concerns by using silicon carbide traction inverters among other features.

Silicon carbide MOSFETs are tested to ensure they meet technical performance and specifications at wafer level before the devices are singulated and then again, in package or multichip module form before they're shipped to customers. However, the extrinsic failure rate or the early life failure rate of silicon carbide is much too high for mission-critical applications such as attraction inverters or even the onboard chargers of the EVs. As such, all silicon carbide MOSFET suppliers apply what is known as a stress test or burn-in test to every device that induces early life failures to happen within hours rather than years to weed out the devices that would otherwise fail in the vehicles.

We are currently engaged in discussions with most other current and future silicon carbide suppliers. The major silicon carbide companies expect that most EV traction inverters will move to multichip modules. As such, they have told us that they must move to wafer level stress and burn-in to remove the extrinsic failures before they put these known good die into multi-die modules to meet their cost, yield and reliability goals of these modules. Aehr's unique, low-cost multi-wafer level test and burn-in solution provides the test electronics and the device contactor technology that enables contact to 100% of all devices on a single wafer and the handling and alignment equipment to provide a total turnkey single vendor solution to meet the needed critical test and stress requirements.

Our benchmarks and evaluations with prospective new silicon carbide customers continue with very good momentum. We've recently completed a wafer stress benchmark with yet another of a large -- of the current large suppliers of silicon carbide with excellent results. They have told us that the FOX platform is the only solution that can scale to meet the production capacity needed to address the silicon carbide device growth, particularly for electric vehicle applications. This is in addition to our previously announced engagement with another large silicon carbide supplier with whom we've been working closely with over the last year to correlate and qualify the FOX system to displace their current production reliability screening test and burn-in systems. The results of that benchmark also met a key milestone this last quarter, and we believe that we will successfully complete their correlation process over the coming months, which will allow them to move forward with our FOX solution. We expect both of these companies to implement the FOX platform solution into their manufacturing production flow.

In addition to the benchmarks with these 2 large silicon carbide companies, we have been approached by several more silicon carbide suppliers to evaluate our FOX-XP systems to meet their production needs for traction inverters and onboard chargers for electric vehicles and also for other applications such as electric commuter train engine controllers, photovoltaic power conversion and other industrial applications. As a result of all these positive evaluations, we believe that we will receive orders from at least several new silicon carbide customers and begin shipping systems to meet their production capacity by the end of our current fiscal year that ends May 31, 2023. With major production releases and ramps of many new electric vehicles from every automotive supplier in the world and many new electric vehicle-focused players coming into the market in 2024 and 2025, there is a significant industry ramp needed to expand silicon carbide production to meet the forecasted needs of these electric vehicles over the next few years and through the end of the decade and beyond.

Now moving to other markets. We're seeing a continued recovery and strengthening in several key wafer level test and burn-in market segments after the last 2 years of softness related to COVID-19. This includes silicon photonics devices for data center and 5G infrastructure, 2D and 3D sensors for mobile and wearable devices and a new high-volume application for data storage on the horizon. Specifically for silicon photonics, during the COVID-19 shutdowns, the data centers such as Facebook, Google and Amazon did not upgrade their data centers from copper-based lands to fiber optic communication links as originally planned. The silicon photonics market had been forecasted to have 30% to 40% cumulative average growth rates for the last few years and through to the end of the decade. However, they end up being flat over the last 2 years with COVID-19 with no growth at all.

We're fortunate enough to work with the market leader in several other key players in the space who all qualify our solution in 2019. However, if they're not growing, they're not buying test systems to meet their increased needs. At the Imec Technology Forum last week, silicon photonics was described as an industrial reality in the critical path of data centers and AI scaling and that advanced CMOS processing and heterogeneous integration will be necessary to address the silicon photonics scaling challenges. We have been told by several of our customers, including our lead customer that wafer level burn-in and stabilization plays a key role in enabling silicon photonics mass production.

Our lead silicon photonics customer that is one of the world's largest semiconductor manufacturers continues to use Aehr for wafer level burn-in and stabilization of their silicon photonics wafers. During the last year, they added a significant number of additional FOX-NP systems to support the characterization and product qualification of new photonics-based devices. This customer is expected to purchase new sets of WaferPaks to be used with these systems. And as the applications and market for silicon photonics-based devices continues to grow, we expect this customer as well as our other customers in this space to continue to increase their capacity in the future.

In just the last month, we received WaferPak orders for new devices from a couple of our silicon photonics customers, and we're expecting customers to resume buying in the current fiscal 2023 and 2024. Several customers addressing the silicon photonics market have forecasted additional FOX systems and WaferPak or DiePak contractor capacity needs over the next 12 months. We expect to see a nice recovery in this market segment sometime over the next year or 2 based on what we're being told by our customers.

We continue to see new programs for our FOX-XP solution for 2D and 3D optical sensors. Last year, we saw yet another device for a new application that we feel will drive our consumables business and possibly require incremental system capacity this fiscal year. We continue to be optimistic that this market segment has significant potential over time, and we continue to meet our lead customer and their subcons needs and to play an important role in their test and reliability supply chain.

Our newly customer in a new very high volume application for data storage devices that purchased the FOX-CP single wafer production test and burn-in system essentially went dormant during the COVID-19 shutdowns. They have begun to show signs of recovery and restarting their planned production capacity rent, which feels like it will begin later this fiscal year or next fiscal year at the latest. We continue to believe that this will drive a significant number of FOX-CP systems.

Now let me spend a few minutes talking about our R&D investments in manufacturing and supply chain. We continue to make investments in our FOX full wafer and singulated die test and burn-in solutions. Last year, we shifted resources away from a planned new package part test system to focus on our FOX products. This year, Aehr will be releasing several test system enhancements that will extend our market leadership of our FOX products for full wafer test and burn-in. These include added voltage ranges, increased parallelism per wafer, new burn-in and stress conditions and a new fully automated FOX WaferPak Aligner configured to fully integrate with our FOX-XP multi-wafer systems to enable hands-free operation. We believe that this will become more important over time for a widespread adoption of wafer level burn-in for multiple markets beyond the markets we address today.

As I've noted before, despite a few bumps in the road, our supply chain is holding up extremely well to the increase in demand and growth, and we've been able to maintain reasonable lead times to meet customer requests. We're very confident in our ability to meet the customer forecasted demand plus considerable upside. I also want to emphasize that we purchased additional material and have a supply chain in place to significantly grow beyond our revenue guidance for the fiscal year. We will have better visibility in the second half of the fiscal year on exactly what that looks like. And once we get closer to understanding the actual capacity needs and requests of our customers will provide an update.

We're very encouraged by the positive momentum we're seeing with current and prospective customers and anticipate multiple new customers will begin placing orders and taking shipments to meet the enormous needs of silicon carbide devices used in electric vehicle market over the next decade. We also see a recovery beginning this year in other key market segments, including silicon photonics and 2D and 3D sensors and another new market opportunity for data storage on the horizon. If current and/or new customers increase our forecast and are designed to pull in orders, we have significant upside capacity to meet their needs. This provides us with the confidence that we can meet a significant upside in revenue shipments that the customer demand pulls in.

We believe we will add several new silicon carbide customers that will be ramping into production by the end of our fiscal 2023 that ends next May. This is in addition to the significant additional investment and capacity by our currently customer for our silicon carbide wafer level burn-in solution. For the fiscal year ending May 31, 2023, Aehr expects total revenue to be at least $60 million to $70 million with strong profit margins similar to last fiscal year. Aehr also expects bookings to grow faster than revenues in fiscal 2023 as the ramp in demand for silicon carbide and electric vehicle -- vehicles increases exponentially throughout the decade.

With that, let me turn it over to Ken to review our financial results and guidance in more detail before we open up the line for questions.

Thank you, Gayn, and good afternoon, everyone. As Gayn noted, we're pleased to report record revenue for both the fourth quarter and full fiscal year. Our fiscal 2022 revenues of $50.8 million were more than 3x last year's annual revenue. In addition to record revenue, we finished the year with record bookings and strong growth in our profit margin. We also finished the year with a solid balance sheet with cash of over $31 million and working capital of $49 million.

Looking at our financial results in more detail. Fourth quarter net sales were $20.3 million, up 33% sequentially from $15.3 million in the preceding third quarter and up 166% from $7.6 million in the fourth quarter of the previous year. These record Q4 revenues reflect our capacity to increase revenues. We actually shipped over $10 million for revenue in the single month of May, which really shows our ability to scale and meet customer demand even in the near term. WaferPak and DiePak revenues comprised 45% or $9.2 million of our total revenue in the fourth quarter. This is our second consecutive quarter of record WaferPak, DiePak shipments, reflecting the growth in the consumables piece of our business.

Non-GAAP net income for the fourth quarter was $6.5 million or $0.23 per diluted share, which excludes the impact of stock-based compensation. This compares to non-GAAP net income of $4.1 million or $0.14 per diluted share in the preceding third quarter, which excludes the impact of stock-based compensation and a $1 million onetime charge for excess and obsolete inventory and non-GAAP net income of $930,000 or $0.04 per diluted share in the fourth quarter of fiscal 2021, which excludes the impact of stock-based compensation. On a GAAP basis, net income for the fourth quarter was $5.8 million or $0.20 per diluted share compared to GAAP net income of $2.2 million or $0.08 per diluted share in the preceding third quarter and GAAP net income of $567,000 or $0.02 per diluted share in the fourth quarter of the previous year.

Gross profit in the fourth quarter was $10.5 million or 52% of sales compared to gross profit of $6.4 million or 42% of sales in the preceding third quarter and gross profit of $3.5 million or 46% of sales in the fourth quarter of the previous year. During the preceding third quarter, the company recognized a charge of $1 million related to reserves for excess and obsolete inventories on legacy parts, which represented a 6.7 percentage point impact on third quarter gross margins. Excluding the impact of this charge, gross margin in Q3 was $7.4 million or 49% of sales.

The increase in gross margin from both the preceding third quarter and Q4 of last year is primarily due to a decrease in unabsorbed overhead costs to cost of goods sold related to higher revenue levels in Q4. Because our manufacturing overhead costs are relatively fixed relative to revenue levels, our gross margins increased significantly with increasing revenues where our fixed costs are basically spread over the larger revenues. As Gayn noted, with the higher revenue we are generating, we are seeing the significant leverage in our operating model to our bottom line as evidenced by the strong growth in gross profit.

Operating expenses in the fourth quarter were $4.6 million, an increase of $507,000 or 12% from $4.1 million in the preceding third quarter and up $1.7 million or 58% from $2.9 million in the fourth quarter last year. SG&A in the fourth quarter was $3 million, an increase of $381,000 from $2.6 million in the preceding third quarter and up $1.1 million from $1.9 million in the prior year fourth quarter. The increase in SG&A expense from the preceding third quarter included an increase in employment costs of $275,000, primarily due to higher incentive payments related to bonuses for exceeding revenue and profitability targets.

The increase from prior year fourth quarter included an increase in employment costs of $768,000. The increase in employment costs included an increase in headcount, salary increases for employees during fiscal 2022, higher commissions and incentive payments related to bookings, revenues and profitability and stock compensation costs related to stock bonuses and our employee stock purchase plan. In addition to the increase in employment costs, the company recognized increases in travel and entertainment and shareholder relations costs.

R&D in the fourth quarter was $1.7 million, up $126,000 compared to $1.5 million in the preceding third quarter and up $626,000 from $1 million in the fourth quarter of the prior year. The increase in R&D from the preceding third quarter includes an increase in employment costs of $198,000 due to higher incentive payments related to bonus objectives. This was partially offset by a decrease in professional consulting of $78,000 as Q3 '22 included milestone payments related to R&D program initiatives during fiscal 2022.

The increase in the prior year fourth quarter included an increase in employment costs of $618,000. This increase included an increase in headcount, salary increases for employees during fiscal 2022, higher incentive payments related to bonuses for exceeding revenue and profitability targets and stock compensation costs related to stock bonuses and our employee stock purchase plan. We continue to invest in R&D to enhance our existing market-leading products and to introduce new products and maintain our competitive advantages and expand our applications and addressable markets.

Now turning to the results for our full fiscal year. Net sales for fiscal 2022 were a record $50.8 million, up 206% from net sales of $16.6 million in fiscal 2021. For the full fiscal 2022, system revenues accounted for 50% of total revenues compared to 44% in fiscal 2021. WaferPak and DiePak consumable revenues accounted for 45% of total revenues in 2022 compared to 35% of revenues in fiscal 2021. Customer service revenues accounted for 5% of revenues in fiscal 2022 compared to 21% of revenues in fiscal 2021.

Non-GAAP net income for fiscal 2022 was $11.7 million or $0.42 per diluted share, which exclude the impact of stock-based compensation, a $1 million adjustment taken in the third quarter for excess and obsolete inventory and forgiveness of the $1.7 million paycheck protection program loan received in fiscal 2020. This compares to non-GAAP net loss of $3.2 million or $0.13 per diluted share, which excludes the impact of stock-based compensation and a noncash net gain of $2.2 million and tax benefit of $215,000 related to the closure of Aehr's Japan subsidiary in the first quarter. On a GAAP basis, net income for the fiscal year was $9.5 million or $0.34 per diluted share. This compares to GAAP net loss of $2 million or $0.09 per diluted share in fiscal 2021.

Gross profit for fiscal 2022 was $23.7 million or 47% of net sales. Excluding the impact of the $1 million excess and obsolescence provision in Q3, gross margin for fiscal 2022 was 49%. This is up from gross profit of $6 million or 36% of net sales in fiscal 2021. Excluding the impact of the onetime charge, the increase in gross margin percentage in fiscal 2022 compared to 2021 is primarily due to a decrease in unabsorbed overhead costs to cost of sales related to higher revenue levels in fiscal 2022. Operating expenses in fiscal 2022 were $15.9 million.

SG&A was $10 million in fiscal 2022, up from $6.6 million in fiscal 2021. The increase in SG&A includes an increase in employment costs of $2.6 million, resulting from the elimination of cost reduction initiatives implemented in fiscal 2021, higher commissions and incentive payments related to increased bookings, revenues and profitability, stock compensation costs related to stock bonuses and our employee stock purchase plan and an increase in headcount. In addition to the increase in employment costs, the company recognized increases in travel and entertainment, shareholder relations and consulting costs.

R&D expenses were $5.8 million in fiscal 2022, up from $3.7 million in fiscal 2021. The increase in R&D includes an increase in employment costs of $1.7 million, professional consulting of $331,000 and project materials of $155,000. The increase in employment costs included an increase in R&D headcount, salary increases for employees during fiscal 2022, higher incentive payments related to bonuses for exceeding revenue and profitability targets and stock compensation costs related to stock bonuses and our employee stock purchase plan. The increase in headcount, consulting costs and project materials is related to R&D programs initiatives during fiscal 2022.

Turning to the balance sheet for the fourth quarter. Our cash and cash equivalents were $31.5 million at May 31, 2022, down $536,000 from $32 million at the end of the preceding quarter and up $26.9 million from $4.6 million at the end of the fourth quarter of fiscal 2021. The increase from fiscal 2021 includes $24 million in net proceeds from our successful ATM offering in the second quarter of fiscal 2022. Cash receivable at quarter end was $12.9 million, up from $8.5 million at the preceding quarter end due to the impact of higher revenue levels. Inventories at May 31 were $15 million, an increase of $899,000 from the preceding quarter end and up $6.2 million from Q4 '21 to support our expected fiscal 2023 growth. As Gayn indicated, we have been ordering long lead components for systems and WaferPaks to ensure adequate supply to meet customer lead times and forecasts.

Property and equipment was $1.2 million compared to $776,000 at the preceding quarter end. Customer deposits and deferred revenue short-term and long term were $2.5 million, a decrease of $3.8 million from the preceding quarter end and an increase of $2.2 million from Q4 '21 related to the changes in our backlog from prior quarters. The company has no debt. This compares to our May 31, 2021 fiscal year-end, where we had $1.4 million outstanding on our line of credit and $1.7 million outstanding on our Paycheck Protection Program loan. Bookings in the fourth quarter were $4.4 million. Backlog as of May 31 was $11.1 million compared to $26.9 million at the end of the preceding third quarter and $1.6 million at the end of the fourth quarter last year. Effective backlog, which includes backlog as of May 31 and all orders since the end of the fourth quarter is $25.5 million.

Now turning to our outlook for fiscal -- the coming fiscal year. For our fiscal 2023 year ending May 31, 2023, we expect full year total revenue to be at least $60 million to $70 million with a strong profit margin similar to last year. We also expect bookings to grow faster than revenues in fiscal 2023 as the ramp in demand for silicon carbide and electric vehicles increases exponentially throughout the decade. Lastly, looking at the Investor Relations calendar, Aehr Test will be meeting with investors virtually at the Needham semiconductor and semi-cap one-on-one conference on August 25. We hope to see some of you virtually at the conference.

This concludes our prepared remarks. We're now ready to take your questions. Operator, please go ahead.

(Operator Instructions) And our first question will come from Christian Schwab with Craig-Hallum Capital Group.

Congratulations on a strong start to the -- strong ending of the fiscal year and outlook for next. Gayn, when you guys are talking about several more customer orders, can you help us understand, is this going to -- should we be assuming that this is going to come from kind of the major European and American silicon carbide manufacturers like STM or Wolfspeed, Infineon, [ROM] et cetera? Or is there an opportunity to get into some of the newer Chinese manufacturers that are ramping as well.

So even in our prepared remarks, we have talked about, I think, starting a few quarters ago, we talked about a benchmark with one of the other major suppliers of silicon carbide today that I think even in February, so we were saying that it was taking longer than we had originally expected. That benchmark, we had a really key milestone that was achieved over the last quarter that bodes really well. We think that we can complete the correlation results, et cetera, over the coming months, and we are expecting to get orders from them.

We also teed up last quarter that we had indications from another -- another of the large customers that wanted to do a on-wafer benchmark. We were actually able to complete that. We brought up the WaferPak the whole thing, and we presented them with the correlation and test results recently and that was -- it was excellent results. And we're pleased with how that's going and think that we can advance that along further. And those were 2 of the large silicon carbide players today, new players that are not current customers. And I believe that we'll be part of their production plans, and that's sort of how I'm describing it. We believe that they'll be going forward with us.

There are others. In fact, there is reviewing the funnel with Vernon and the team there is a significant number of players that are talking about getting in the silicon carbide. Many of them have been announced that include large and small. And candidly, many -- I would actually say, large players that have not announced their plans that have approached us and have been talking to us about their very real plans for engine in silicon carbide market. Now folks like Canaccord Genuity have actually been ringing the bell for the last really 6 months, pointing out that the current announced plans by all the major silicon carbide suppliers only supplies about half of just the electric vehicle needs by the end of the decade. So it's welcome to have all these new players coming in because all of the players today don't have a chance of meeting all the demand as they basically go from 1x to 25x the output by the end of the decade.

So we're leaving it kind of open, not trying to be too vague here, but there's -- we're going to have a pretty significant number of silicon carbide customers, and we're just trying to make sure people understand that, that will include at least several this year that will not only buy from us, but we'll already start taking production orders for their ramps.

And then, Gayn, when we talk about wafer starts in silicon carbide for the electrification of the automotive industry. Do you -- have you guys -- I know yields are substantially different potentially by -- by different customers. But for every -- can you give us an idea if you've been able to finalize [you] not have enough data for every 25,000 or 50,000 wafer starts if that was -- if it was a customer using you, how much capital equipment they would need to buy. And the reason why I say that is the wafer start numbers between Canaccord and others out there in the industry such as Yole and such as the lead provider on the capital equipment side for an implant to really doesn't face much competition. Those -- the trajectory of growth is very similar. The exact wafer starts aren't all that close. So I'm curious your thoughts on that.

Yes. Okay. So let me try and answer it with a couple of different ways. So first of all, like we get a front row seat of people's yield, and we sure as heck don't want to talk about it publicly. So when you look at the Canaccord's of the world, et cetera, they're taking their assumptions of yield into account. I think most people understand, and it's widely known that the yield, the negative yield of silicon carbide devices is not extremely high, let's call it that. But it's also not -- it's not crazy low either. So I'm just not going to answer that. But specifically, if you look at gross die per wafer, which is prior to yielding, that data is out there. There's a lot of data points out there because people like Wolfspeed and others sell that singulated die, so you can -- they actually describe the size of the die.

And the numbers that we can see out there is for an inverter for an electric vehicle. They tend to be about maybe 350 to 400 and some devices, gross die per wafer at 150 millimeter today. And then when you go to 200-millimeter that's going to start kicking in towards the end of the decade, in particular, I mean, I don't think the midpoint, 50% till late in the decade, split between 200 millimeter, 150, 200 millimeter is about 75% more die per wafer. We've been using the number of approximately, and this is averaged over this whole period. Like maybe 500 die per wafer is all you're going to get off of these wafers for, call it, 6-inch equivalent wafers. And that's rounding up. I mean -- and then we know that there's about 48 die or there are 48 die per inverter, which is per engine. And so I've seen people -- some of the silicon carbide suppliers use numbers like 4 engines per wafer to maybe 8 or 10 on the high-end side.

And so you can start looking at each -- how many wafers are needed to supply a single engine car. The Model 3 is what 70%, 80% of their cars are sold with 2 engines, et cetera, et cetera. The current, believe, I think, is on the 1.5-plus average engines per car at some point in time. So we can see where the demand is coming from. From our tool, it really comes down to the test and burn-in time, and that is something that continues to be dynamic. We have talked publicly about -- there have been public statements by some of the suppliers out there that have talked about 6 hours. They've talked about 14 hours. They've talked about 24 hours. In some of our modeling, we've talked about test times being able to get down to, say, 12 hours of burn-in and on a system like ours with 18 wafers that theoretically means you could get 36 wafers per day out of it.

Now we know for a fact that there's a number of conversations going on in the industry that the burn-in times are likely to go up there's as much pressure to go up as to go down. And the pressure is that there still is more opportunity to increase the quality by increasing the burn-in times to remove infant mortality to get down into the sub-100 or even 10 part per million numbers. There's a lot of capacity today that is being shipped in die form. And there were public statements that talked about being able to ship the die form with 99% of the failure screened out. Well, that's awful. I mean, it means you have a 1% of the die still need to go through some sort of quality burden at a later date. There's actually 2 big suppliers out there that supply die sales today, and one of them was the one that actually talked about that 1% failure rate. That's just not going to be good enough. We think that you need to be down in the 100 or single-digit part per million failure rates and you're going to need to have longer burden times.

Lastly is folks like Danfoss and BorgWarner and companies that are only buying singulated die are demanding and specifying stabilized threshold voltages of the devices before they put them into the multichip modules. And that actually takes a considerable amount of time to actually stabilize that through a burden process. So those 2 things are driving test times up. There's obviously cost pressures to try and drive it down. And we think that test times will continue to be in the on average multiple hours up to 12 or perhaps even 24 hours depending on the customer and where the product is throughout the decade.

And then my last question, again, kind of on the wafer start side, others are projecting that silicon wafers for the automobile industry will have to at least triple over a 3-year time frame to potentially hit the penetration rates that are expected. So as we look at your leading customer and let's assume that they grow with the industry and their market share remains whatever it is today. I guess there's no reason to believe that they couldn't be at least 3x larger than they've been over the last 12, 14 months of orders. Is that the right way to be thinking about that?

Yes. I mean it's always awkward. So we announced ON Semiconductor as a 10% customer last year. ON Semiconductor was a distant sixth or seventh place player in a 3-man race 2, 3 years ago. I mean, it was -- they were -- they did a couple of few million dollars worth. It's very clear that they are taking a significant amount of share and they're boldly out making comments about the first company to reach $1 billion in silicon carbide, this again from a company that a couple of years ago didn't do $10 million. They also just announced $1 billion commitment to $1 billion. It was in Korean won but into their Korean facility for silicon carbide. And so they're making some significant investments, and it's going to take them and a whole lot of other people to supply the demand.

I listened, I listened to ST and they were talking about the capacity they're increasing in Singapore over the next year and then a new fab coming online in Sicily, Italy in 2024. And they were saying they don't believe that they're going to be able to sustain their market share even with adding an entirely new fab. So [they're going] to take a lot of new players, and there's -- we kind of look at it as a land grab. There is a lot of folks getting into here. We're trying to run as fast as we can. We're adding resources in sales and marketing and other things to try and get -- make sure we get our message out there. We're adding applications engineers and doing some things structurally to be able to address more customers because there's going to be a lot of new and big players out there from our perspective. So pretty exciting time.

Our next question will come from Larry Chlebina with Chlebina Capital Management.

I got a quick question on your anticipated additional sales, I think you said in automating the XPs going forward. Is that -- last August, you had a sale to your lead silicon carbide customer for an automated aligner. And is that going to revenue anytime soon? Or did you already ship...

Yes, it is going to revenue anytime soon. So you're actually bouncing around a different question. I'm just going to share it here. We -- that order, we talked to the customer about it, and we have converted that order into our next-generation automated aligner. That new liner is available in both what we call a stand-alone or an integrated form. Sometimes people want to do as we do today with our volume customers, which is they share an aligner across multiple systems. There's other customers that want to take the aligner and bolt it right onto the front of it and remove the people that are actually moving the WaferPaks around similar, by the way, to how all packaged part burn-in has done today is with people moving around these bids. In our case, they're called WaferPaks.

So -- and in fact, the order that we announced today is the second order for that system. Those aligners will be shipping over the -- I mean, the second one will actually ship before the end of February as well. The automated version of it is just more integrated. It has some additional R&D that we're going to be working on over the next 3 to 6 months. And we are now -- have quotations out and we'll be accepting orders for it in the automated form for delivery sometime towards the end of our fiscal year.

So that -- the one you, you put in the press release today from -- for the $12.8 million. That's actually the second automated aligner.

Yes.

And so the first one -- the first one you expect revenue, when did you say again? Did you mention that?

Yes. I actually haven't given a commitment on that. But prior to the second one, which is revenue in February, how is that...

I hope so. So since this is your lead silicon carbide customer, are they -- do they have intentions of automating all of the XPs that they eventually either have or going to receive? Is that kind of what your expectations are?

Not at this time, no. They like other customers are totally convinced they want to be off-line, because of their extended burn-in times and the way they use the tools. And so we offer the ability -- we kind of say, yes, to every customer. So this gives us an opportunity to -- there our customers that absolutely feel strongly one way or the other, and we want to be able to meet their needs.

So some of the potential new silicon carbide customers may be fully automated. Is that...

I believe that is true. That's correct. Yes.

Okay. Then initially, this application or a fully automated XP. By the way, what are you calling that thing? Do you have a name for it yet?

We haven't named that system yet. We'll get a name out later this year.

Okay. So the initial application, I thought was for memory, specifically Flash Memory, for stacked eye applications in order to get rid of the infant mortality risk on one of those stacked eyes. Is that -- do you expect to kick that off anytime soon now that you're on a verge of launching this new product yet to be named new fully automated XP?

So it's always been a little bit hidden down in the comment sections, but we've teased people at every quarter along the way. We had actually done some design reviews with a couple of key memory suppliers on this aligner right before COVID. And all of those guys went very, very quiet. During COVID, we've had conversations that have been, quite frankly, a little bit in the background. Their comments to us were at this point, they're ramping what they have. Nobody is doing evaluations, which is, I think, has been the experience across the whole semiconductor test business. Nobody was really buying new things. They were buying what they had before. So we've had some recent conversations that lead me to believe that there's the opportunity to renew those discussions is upon us.

With a slowdown in memory slightly that they would be more open to kick some off like this. But the question is, is it your intention or do you think the best way to look at this would it be an existing fab that would be adding this capability because it would be so-called...

No, definitely not. My experience is -- yes, my experience is usually the best time to cut in is at a discontinuity such as a new fab floor going in. And so...

Even though there are cost benefit might be so great that it could pay for itself saying 6 months or...

That has been my experience, in particular, as they add these new fabs, and they will put a floor space in plan consistent with whatever it is that they plan to put the tooling in on, that would be my experience.

Lastly, along those lines, with that new capability, would it make sense the joint venture with maybe an ATE testing company to incorporate since it's a fairly long burn-in where maybe they could do the ATE test on a wafer. While you're burning in 18 wafers, they could do maybe one at a time, but get 18 done during the entire burn-in cycle. Is that something that you guys might be considering because that seems to be to make a lot of sense?

We're getting a lot in the futures here, but let me just make a plug. There's not an ATE product out there that has the level of density and power management that could fit into our XP. That's one of the key differentiations that we have is that we're able to put a 2,048 channel ATE system into basically a 3.4-inch pitch and put 18 of them in the same footprint as a regular ATE system with not dissimilar pin count testing one wafer. So, I would not -- if I could, I couldn't pick one ATE supplier out there that I wish they could build a tester and stick into my system. I think we're actually more differentiated.

The FOX-XP system and its base hardware was in fact, defined around capabilities that are much more complex, certainly than a MOSFET, for example. It has full digital capabilities, pattern generation, et cetera, and it is capable of testing these flash memories. So, I actually I'm pretty proud of the system. We spent a lot -- we haven't talked a lot lately. Now there are other ways to potentially partner with some of the supply or some of the customers who themselves have done some of their own work with test and that could be something to think about. But most days I wake up and I focus my energy on what's going on exciting in the silicon carbide space. So...

Right, right. Well, looking down the road...

But I'm committed to the memory business in time as everybody who has heard me talk knows.

All right. That's all I had. Thanks. Again, great quarter and keep up the good work, good year, look for a better year.

Thanks, Larry.

Our next question will come from Dylan Patel with SemiAnalysis.

I wanted to ask about this test intensity sort of question. So, many other test firms sort of speak about the relationship of lower yields, meaning higher test intensity. And right now, silicon carbide deals aren't pretty even at the best firms, but they're expected to go up. My understanding of your solution is that it's unique in that because it's burning, you basically test the same regardless of how lower high yields are. Is that accurate? Or do -- I get that loan yields necessitate, people would want to buy your product. But can you expand on the durability of demand even if yields reach multiple 9s eventually?

Yes. Okay. So, the -- there is certainly correlation between low-yielding devices and devices that are more apt to need reliability and burn-in. But at the same time, the failure mechanisms are not the same. The primary mechanisms that cause the yield and the devices are not the ones that are necessarily induced by the burn-in. The burn-in sometimes require additional energy through heat and voltage and power to actually create the tunneling effect or negate that causes this failure. You can't actually test for it. You have to actually do a stress test to find it.

So, there are devices with low yield that don't use burn-in, the application doesn't need it. There are devices with very high yield that need 100% burn-in. DRAM would be an example of it. DRAM has extremely high yield, but it still has 100% burn-in to catch about a 1% defect and has now for 40 years. So, that's a scenario where there's way more money in it to get rid of burn-in in DRAM than even silicon carbide and no one has figured out how to do it. The information that I'm hearing from the silicon carbide folks is that they don't believe that there's any chance that burn-in is going to go away basically in the horizon of a decade or more.

One thing is going to happen is there will be ways to try and optimize test time by increasing temperatures and doing things to try and optimize, but you're always riding the line that if you stress the device a little too much, you can actually damage it. And so there's a balance. One of the analogies I've used is microwave versus an oven. You can actually cook a turkey in a microwave, but it doesn't come out that well. And similarly, if you apply enough energy to a silicon carbide, you can induce the failure within potentially seconds, but you will absolutely damage it and it won't last for months, much less years.

So, I've actually found my experience so far consistently with the customers I've been working with. They're actually applying less thermals and less voltages than I thought they would because they're trying to keep their devices to have long-term or intrinsic reliability to meet the needs. So, there's always sort of a balance there. Again, I don't believe that I think as yields go up -- one thing I actually heard one of our customers talk about is that the next generation and generation after, the device yield is actually expected to improve, but the burn-in times are going to get longer, which is definitely not intuitive. And the reason they said is it's going to take longer to actually get the failure. It will take longer to induce the tunneling effect that creates the failure. But if they don't, then the part instead of it dying 3 months later in a car, it would dive maybe 2 years later in the car. And that's catastrophic if your entire fleet of cars were to start failing with walk-home events 3 years later. So anyhow, there's a lot of science involved in this thing, and it does appear at this point that burn-in is here to stay for a while on these devices.

Thank you for the question.

And then I wanted to switch gears to wafer packs related to silicon photonics. Specifically, where is the burn-in specifically happening on the silicon photonics? Is it happening on the indium phosphide wafer where the lasers were fabricated or is it happening once you bond them onto the silicon photonics wafer?

Well, we have examples where customers want to burn it in on a wafer, call it, a substrate that would then be bonded on. But the examples of our current customers are all in a scenario where -- well, on the wafer level customers where they're bonded on and then they are burnt in after they're bonded on. So -- and customers like Scorpios, which we've announced, they actually fabricated on. It's printed on, if you will, as a process node. So, it's not really a bonding, if you will. And they do the burn-in on the wafer level for scale before that die is singulated and then placed into the system.

That was helpful for trying to understand when there's new generations of lasers or if it's when there's a new, say, transceiver and it's the latter. And just one last little question is in the past, we've talked about difficulties visiting your major prospective clients, especially in Asia, like China, South Korea. I've luckily been able to start visiting some of my clients in Asia. Have you been able to? And do you think that's going to help move the needle on some of these future customer orders or evaluations?

It is. I mean Europe has been pretty free and getting better. You can even get back to the U.S. without passing a COVID test as of like last week, but we've been able to freely move there with some, call it, some restrictions. We've had people that are moving to and from the U.S. and Asia. And we've got some -- we have a recent Asia customer that's been making visits here. So, I think it's opening up. China is still pretty iffy. But as that opens up, we think that will be better. Yes.

Our next question will come from Bradford Ferguson with Halter Ferguson Financial.

My clients are now 1% holder of your company now, so shareholders. I was curious what's keeping the likes of these the serious silicon carbide makers from creating their own wafer-level burn-in. And curious what's protected? Is it the 18 wafers at a time? Is it the aligner or the loaders? What's protected from that?

There's actually a lot of protection. Is it Brad? I guess, it's nice meeting you. So -- wow, our systems have a significant number of patents and IP that make a very significant hurdle to get into. If you understand our space, an expert in the test space, they will tell you that being able to test a wafer is not novel. Every semiconductor wafer in the world has been tested and they are all tested in a similar way with what's called a waver prober, a probe card and a tester that sits on the wafer probe. I built them for years.

When I built a tester, I had to build a tester that would always sit on the generic platform of the 3 major prober suppliers and each of their 5 or 6 probers. So, there's like maybe 20 probers back on or so, maybe it's less than that, that the tester was generic to. The probe cards all had to be worked with. If you want to build a tester, you had to build it with -- that would work across half a dozen to a dozen probe card suppliers. And all the folks like form factors and Technoprobe, et cetera, have standard design packages for a specific tester and the product teams and the tester groups work with them to ensure that the probe cards are there when the testers come out.

Similarly, the docking to the prober. These sort of 3 pieces create through a commoditized and stable market because you could buy any one of the 3 probers, any one of the 20 probe cards, any one of the 5 or 6 testers out there. And that's how the world has operated. With us, the problem was, you couldn't break through a cost barrier, if you start with a $300,000 prober, that's the footprint size of a Prius, in a parking lot. And you need to put 100 or 200 or 300 wafers worth of capacity in place because you have a 10-hour burn-in time, you can't get there. You just -- the cost of the depreciation of the capital equipment sitting in a wafer test environment is too high.

So, what we have is we've created a prober-less tester. It actually doesn't use a prober. It doesn't use a standard probe card with the normal force displacement. It's planarity comes completely different than the patents in the wafer -- in the probe card industry. And we use a tester with a tester interface that doesn't exist by anyone else. There are several features about our system that literally do not exist by anyone. So, if a tester supplier want to come in, first of all, they would step all over our patents, everything from how we contact a wafer pack to how we distribute power to onboard components. There's a number of patents that we do that they can't do, this blade architecture.

If you're a probe card supplier, you would have to build a probe card that would look just like our wafer pack that absolutely positively would violate a half a dozen or more of our patents. We've approached some of the big suppliers about potentially being partial suppliers to us. And some of those big guys, their probe tips would rip completely off the wafer pack because we run 150 C and we run them from room to 150 C and back without a soak-time. And those proviant -- the men's-based probes would never survive. Okay? We have direct quotes from people on that. So we do it differently than them.

And the prober companies have no vested interest in being involved in an a prober-less prober -- in a prober-less tester. So in fact, that was part of the real risk associated with us for the early investors. In order for somebody to actually go with our solution, they had to close their eyes and have confidence that we would supply with them the tester, the pro card and the handling equipment because if any one of those didn't work, nothing would work. And so there was a point in time where it was like, is anyone going to be -- have the nerve to actually do this.

And then looking back, as people know now, our lead customers that became 10% customers included the likes of Intel and Apple. And now it looks a lot easier with some large suppliers that are ramping but if you were to go out and try and build what you have, there's no way to do it without violating our patent, and it would cost you a lot of money and a lot of time to try and do it even if you want to violate the patents. So, we have an enormous head start and IP advantage. And right now, we are not aware of any company who has been told that anybody is working on a prober-less system like ours. And we keep our ear to the ground so that we can quickly take legal action against anyone who tries.

I'm guessing that the FOX system can handle 200-nanometer today.

200-millimeter wafer size. Yes, 200-millimeter wafer size is what you meant. Yes, we can. We actually do 4-inch, 6-inch, 8-inch and 12-inch wafers today with our wafer packs. They're all in volume production, and all of our FOX wafer pack of liners can handle all of the above.

So, 12-inch is effectively 300-millimeter, is that right?

That's correct. Yes. Yes.

Okay. So, several of the serious players are creating new fabs that are dedicated to silicon carbide and gallium and they're intending to grow at -- they're intending to grow at 40% a year with their silicon carbide business. Meanwhile, we look at EVs growing at 100% a year. So, it seems like maybe they're being a little conservative and you all are guiding to a high end of a 40% annual revenue growth. What makes you all a little on the conservative side? Is it wondering whether Photonics is going to come through finally this year? Or is it just a nervousness about when orders come?

I think it's just timing of orders and when, not if. We've alluded to our order growth is going to be higher. We've never guided orders. We talked about it as a board, whether we should do that. So, we've at least said for the first time, they're going to be higher. The question is when is the timing happening. I mean we've talked about this in previous calls. There were some of those big silicon carbide companies that 1 year ago, were talking about opening up a fab by the end of the calendar year, like last December. And we -- and people were beating me up and I felt bad that how come you haven't sold them a wafer-level burn-in system. And the reality is, I was questioning that ourselves trying to understand what were we missing till the dust settles and you realize that they're not even sampling out of that fab yet. And the production isn't going to be really until next year.

So, I think that the timing of these new fabs and their output is the piece we're watching. But there is no doubt that some of that time is behind us and the pending orders coming from the automobile companies are driving big dollars to put this capacity in place during '23, definitely '24 to meet the '25 ramp. And so that's where the hesitation comes from. And we're doing a ton of things in the background to our supply chain to ramp it up to be able to do things faster to do. We're doing more outsourcing with our contract manufacturers so that we can assemble and test the system in a much shorter period of time.

And I was -- I mean that data point, folks, we ship a system -- and on the eighth day basically or the ninth day, they tested 18 wafers, okay? It's like shipping 18 testers and having them all released 8 days later. That is crazy and that's awesome. That both talks to how much demand there is and how they're pulling. But Vernon and his team and my support guys are that's just awesome that they could do that. And as we do that, that allows us to ramp faster and faster for people. So, I'm trying to be cautious and appropriate with setting Street's expectations, but I don't think I've ever been more optimistic than we are right now.

And one more question. Revenues versus last quarter grew $5 million, but your COGS only grew $1 million. So, even if you adjust for that inventory write-down that happened it seemed like your COGS only grew by $2 million. So to me, that says that you have potentially a 60% gross margin as you scale bigger and bigger. As you...

You're within a point. You're right, you're dead on. Yes, what we call our direct margin, which is basically our total costs, including warranty, et cetera, are about $0.40 on the dollar. And our margins are actually improving from last year to this year in direct margins. As we go to 200-millimeter wafer packs, our margins are actually improving. But the customers are delighted because they get more -- the lower cost associated with it. Our new systems are actually we're expecting to see improved margins related to some of the new features that we think provide additional value to customers, lot of things are going right for us.

Our next question will come from Wally Wadman with Constitution Research.

A couple just filling in the blanks. Can you hear me, okay?

Yes. Perfect.

Okay. Did you -- maybe I missed it. Did you say what percentage of your revenue is you lead customer was either for the year or for the quarter?

It's 82%.

Yes, it was a significant number at over 80%. 82% is what Ken is telling me across the board right now.

And I'm also curious, when you talked about going forward, your R&D spending more spend -- you focused on the silicon carbide market, now you're focused on, let me call it, the legacy markets. But I'm curious if that spend is going just to upgrade and enhance existing tools? Are you trying to develop new tools or new markets?

Yes. It's -- there's a little of both. There's some things that we're doing to add some voltages and all that some customers have specifically asked us. They would allow us to win that customer, but they're in the same market because remember, everyone doesn't do everything exactly the same. There's also things that we really would consider new markets and opportunities. And I think particularly the automated aligner, there's going to be certain classes of markets that really makes a lot of sense. So, it's a mix, but there's a lot of things that we're doing just to make sure we shore up to capture sort of the markets that we're in, capture a larger, if not dominant market share of that and address all customers. But we're also doing things that we think are going to play forward like, for example, the applicability to memory that Larry was talking about.

And then you've talked over the last couple of quarters about potential new customers. I'm curious of that $50 million annual revenue, fiscal year revenue. How much of that came from existing customers and how much came from new customers, however, you want to answer that question?

Yes. It was pretty -- I mean, we consider our lead silicon carbide customer and existing customer. So, we just talked about 82%. I think when we started the year, we were expecting and I think I would have shared that, that we were going to get 1 or 2 probably single system orders or something by the end of the fiscal year and they didn't come in from the new customers. We're excited about where we're at right now heading into the year, but it's sort of surprising to me that some of that stuff didn't happen as fast as it did. We did not lose those customers. They did not change their mind. They have -- they're more engaged with us than they were a year ago in this case, but it's been a bizarre -- it's sort of taking some time.

This year's -- this year's revenue, I won't give specific numbers, but we believe a good chunk of that is going to be new customers and has a potential some of the growth is from those new customers too. But candidly, our current customers could easily surprise us just based on the tone they're talking about and kind of reengaging. And so we're trying to make sure we have plenty of material and the supply chain to be able to address that mix, if you will, during the first half of calendar year next year, in particular, whether there is more uncertainty.

I don't know if you are willing to answer this enough. But the last couple of quarters, you've talked about being able to handle volumes above what your existing forecast. I'm just curious what your sales capacity would be or what it is? Could you -- if the business materializes, $100 million or $80 million, what does that look like?

Well, I'll tell you what, I'll give you this one. Ken mentioned it the other day. So, we -- just by the math, we did 20 point whatever million in Q4. Over half of that was in May. Okay? So, we're doing $120 million run rate.

I can write down $120 million in my model. No, don't forget about that.

We have proven to ourselves that we have the ability to do that right now. And I don't believe that's enough as we go forward over the next several years.

Last quick question. Pricing, were you just passing through your cost increases? Or were you able to get out in front of that?

You know what, we've been able to manage our vendors fairly well and penalize them drastically when they try and raise their prices with us. I just disdain that. But there have been certain things like shipping -- oh my gosh, last Christmas, we normally would ship chambers. We have 2 chamber suppliers and we ship them. They're built in low-cost regions. They're to our specifications. And normally, our chambers would cost us like $2,000 if we put them on a boat each to ship them to us. We wrote a check for $496,000 for a pair of them to get them here right after Christmas because we -- they couldn't get on a boat, L.A. was locked up. You want to see it on a plane, (inaudible) a piece. It's crazy. We wrote the money. I did not pass that on to the customer. I had a tear in my eye and explain it to them. And those costs are coming back down again. It's our hope to try and manage this. We are seeing some raises of some prices in certain areas, but we've been able to manage it fairly well.

Our next question is a follow-up from Christian Schwab with Craig-Hallum Capital Group.

Just a quick follow-up, Gayn, is there any reason other than, I guess, lack of success by new customers designing new end-to-end customer shipments, but we've already discussed at great length, the rapid growth that's going on. That these other customers couldn't ramp in a 1-year plus or minus time frame from when they start giving you orders to a similar level as on? Is it fair to assume something similar to that?

Yes. I mean, yes and no. I happen to have a pretty good idea of what those guys are. So, it depends on the time line. So, if you said, within a year, can they be as big as on from the first time they take a system, I would say yes. Could they be sooner? Sooner than what, sorry, on our original customer did, probably. Our first customer bought a system and then didn't take another one for, what, 1.5 years. That's not going to play out with the new guys, I don't think. I think they're going to go much faster. Do they have the ability to do $20 million, $30 million of fees? Absolutely in a year.

I show no further questions. I would like to turn the conference back over to management for any closing remarks.

I appreciate it. And thank you, folks. We've had some feedback that our conferences do run aways, but I wanted to make sure, last time we ended up cutting off before everybody got their questions and so I'm glad that people had a chance. So with that, I appreciate everyone's time. We're really excited about this fiscal year, and we look forward to seeing you at one of the investor conferences or on our next call. Take care now. Bye-bye.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.