Aehr Test Systems
NASDAQ:AEHR
US |
Johnson & Johnson
NYSE:JNJ
|
Pharmaceuticals
|
|
US |
Berkshire Hathaway Inc
NYSE:BRK.A
|
Financial Services
|
|
US |
Bank of America Corp
NYSE:BAC
|
Banking
|
|
US |
Mastercard Inc
NYSE:MA
|
Technology
|
|
US |
UnitedHealth Group Inc
NYSE:UNH
|
Health Care
|
|
US |
Exxon Mobil Corp
NYSE:XOM
|
Energy
|
|
US |
Pfizer Inc
NYSE:PFE
|
Pharmaceuticals
|
|
US |
Palantir Technologies Inc
NYSE:PLTR
|
Technology
|
|
US |
Nike Inc
NYSE:NKE
|
Textiles, Apparel & Luxury Goods
|
|
US |
Visa Inc
NYSE:V
|
Technology
|
|
CN |
Alibaba Group Holding Ltd
NYSE:BABA
|
Retail
|
|
US |
3M Co
NYSE:MMM
|
Industrial Conglomerates
|
|
US |
JPMorgan Chase & Co
NYSE:JPM
|
Banking
|
|
US |
Coca-Cola Co
NYSE:KO
|
Beverages
|
|
US |
Walmart Inc
NYSE:WMT
|
Retail
|
|
US |
Verizon Communications Inc
NYSE:VZ
|
Telecommunication
|
Utilize notes to systematically review your investment decisions. By reflecting on past outcomes, you can discern effective strategies and identify those that underperformed. This continuous feedback loop enables you to adapt and refine your approach, optimizing for future success.
Each note serves as a learning point, offering insights into your decision-making processes. Over time, you'll accumulate a personalized database of knowledge, enhancing your ability to make informed decisions quickly and effectively.
With a comprehensive record of your investment history at your fingertips, you can compare current opportunities against past experiences. This not only bolsters your confidence but also ensures that each decision is grounded in a well-documented rationale.
Do you really want to delete this note?
This action cannot be undone.
52 Week Range |
10.17
29.31
|
Price Target |
|
We'll email you a reminder when the closing price reaches USD.
Choose the stock you wish to monitor with a price alert.
Johnson & Johnson
NYSE:JNJ
|
US | |
Berkshire Hathaway Inc
NYSE:BRK.A
|
US | |
Bank of America Corp
NYSE:BAC
|
US | |
Mastercard Inc
NYSE:MA
|
US | |
UnitedHealth Group Inc
NYSE:UNH
|
US | |
Exxon Mobil Corp
NYSE:XOM
|
US | |
Pfizer Inc
NYSE:PFE
|
US | |
Palantir Technologies Inc
NYSE:PLTR
|
US | |
Nike Inc
NYSE:NKE
|
US | |
Visa Inc
NYSE:V
|
US | |
Alibaba Group Holding Ltd
NYSE:BABA
|
CN | |
3M Co
NYSE:MMM
|
US | |
JPMorgan Chase & Co
NYSE:JPM
|
US | |
Coca-Cola Co
NYSE:KO
|
US | |
Walmart Inc
NYSE:WMT
|
US | |
Verizon Communications Inc
NYSE:VZ
|
US |
This alert will be permanently deleted.
Earnings Call Analysis
Summary
Q4-2024
Aehr Test Systems reported record revenue of $66.2 million for fiscal 2024, a 2% increase year-over-year despite a decrease in Q4 revenue. The net income rose to $35.8 million, driven by a significant tax benefit. Despite challenges in the electric vehicle market affecting silicon carbide device orders, Aehr remains confident in this market. The company invested in enhancing its WaferPak products and acquired Incal Technology to bolster its presence in the ultra-high-power semiconductor segment. For fiscal 2025, Aehr projects revenue of at least $70 million and expects silicon carbide and emerging markets like AI processors and gallium nitride to contribute significantly.
Greetings. Welcome to the Aehr Test Systems Fiscal 2024 Fourth Quarter and Full Year Financial Results Call. [Operator Instructions] Please note, this conference is being recorded.
I will now turn the conference over to your host, Jim Byers at MKR Investor Relations. You may begin.
Thank you, operator. Good afternoon, and welcome to Aehr Test Systems' fiscal 2024 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, Chris Siu.
Before I turn the call over to Gayn and Chris, I'd like to cover a few items. This afternoon, right after market closed, Aehr Test issued a press release announcing its fiscal 2024 fourth quarter and full year results. That release is available on the company's website at aehr.com. There were 2 other announcements issued today, and those are also posted to the company's website.
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.
I'd like to remind everyone that on today's call, management will be making forward-looking statements 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 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 now, with that said, I'd like to turn the conference call over to Gayn Erickson, President and CEO. Gayn?
Thanks, Jim. Good afternoon, everyone, and welcome to our fiscal '24 fourth quarter and full year earnings conference call. Thanks for joining us today. I'll start with a quick summary of the highlights of the fiscal fourth quarter and full year we just completed in May and spend some time giving an update on the key market areas addressing for semiconductor wafer level test and burn-in, including some new emerging opportunities. I also want to go over the exciting news we announced today with the acquisition of Incal Technology, which has some incredible products addressing the ultra-high-power semiconductor market, including a significant number of AI processor makers. Then Chris will go over the financials in more detail and provide our guidance for the new fiscal year. After that, we'll open up the lines to take your questions.
Starting with our financial results, as we reported in our prelim announcement last week, our full year revenue and net income results exceeded our previously provided guidance and surpassed analyst consensus. Although we saw customer push-outs of silicon carbide devices due to slower electric vehicle demand in the second half of our fiscal year, we still achieved another record for annual revenue for Aehr of $66.2 million.
On the bottom line, GAAP net income was $33.2 million or $1.12 per share, which includes a tax benefit resulting from the release of the company's full income tax valuation allowance of approximately $20.8 million recognized in the fourth quarter. Chris will talk more about that.
This past year, wafer level test and burn-in of silicon carbide power semiconductors used in electric vehicles, or EVs, were a key driver of our business. And we anticipate that market will continue to be a key contributor to revenue in the current fiscal year. We're also seeing traction with several emerging opportunities for our test and burn-in solutions in new target markets and expect bookings and revenue across a much broader range of customers and markets this fiscal year. These new target markets include quality, reliability and production test and burn-in of artificial intelligence processors; wafer level burn-in of flash memory devices used in solid-state disk drives; burn-in of semiconductors used in hard disk drive magnetic rewrite heads; wafer level burn-in of gallium nitride power semiconductors used in data centers and solar power conversion; and stabilization and burn-in of silicon photonics integrated circuits used for optical I/O communication between chipsets and processors.
I'll cover at least a little on each of these key markets, beginning with wafer level test and burn-in of silicon carbide devices. We continue to have a high level of confidence in this market, which remains an enormous opportunity for Aehr. While most forecasters are saying that the inflection point for silicon carbide and electric vehicles is now the second half of 2025 into '26, from our many meetings with semi suppliers, Tier 1s and electric car companies themselves, it's even more clear now that silicon carbide is the plan of record for electric vehicles and preferred over IGBT. Virtually every car manufacturer is designing new electric vehicles with silicon carbide modules, which absolutely need reliability test and burn-in to screen out failures that otherwise will show up in the life of the vehicle. Burn-in of the die at the wafer level before the modules -- before they're put into modules is significantly more cost-effective with much higher yield than doing this at the module level. We believe we're in a strong position to win more than our fair share of this business as we believe we have the industry-leading wafer level burn-in solution.
This past year, we engaged with a significant number of new silicon carbide device and module suppliers related to their anticipated capacity needs, and we remain engaged with these and all major players in the market, including many in China. We continue to make great progress with our previously announced benchmarks and engagements and believe these potential customers are committed to wafer level burn-in to meet their requirements for good die for die sales and for their power modules.
The silicon carbide market continues to be an enormous opportunity for us, and we're seeing more and more auto suppliers that are committed to silicon carbide in their EVs, as well as road maps that are based on modules for their electric motor power inverters. We're also seeing growing demand for silicon carbide devices beyond the EV market such as solar, data center and other industrial applications for power conversion. We remain very enthusiastic and believe we are well positioned to continue to grow our business in silicon carbide and expect to receive first orders from a significant number of additional silicon carbide customers by the end of this fiscal year.
Today, we announced that we received over $12.7 million in orders from one of our silicon carbide customers for WaferPak full wafer Contactors to be used for production needs for wafer level burn-in and screening of the silicon carbide devices for the EV market. We're excited about our continued partnership with this customer and to receive these orders to help them meet their needs for new device designs.
As these orders illustrate, when our customers win new designs from their customers or they change device designs, wafer patterns or sizes, these customers need to order new WaferPak contactors from Aehr to fulfill these design changes. This consumable type of revenue grew in fiscal 2024 for us, representing 57% of total revenue as systems orders growth slowed, but new designs and variety of devices increased, causing incremental WaferPak sales on the installed base.
As we look ahead, we believe that silicon carbide remains a very large market opportunity for Aehr as more and more EV manufacturers adopt silicon carbide, and we believe we're well positioned to continue to capture market share. We expect to add a significant number of silicon carbide customers both this fiscal year and the next fiscal year as silicon carbide ramps in the second half of 2025 and into 2026.
Now, let me talk about the AI processor market. Last month, we announced we're working with an AI accelerator company to move their AI processor test and burn-in to wafer level and have secured a commitment from them to evaluate our FOX solution for production level test and burn-in of their high-power processors. This company recognizes the potential of the significant benefits of production test and burn-in of their accelerators while still in wafer form before they're integrated into the end application product, which would prove to be more cost-effective and significantly more scalable than doing the screening later in their manufacturing process. We think this is an amazing opportunity to displace the current package and system level test for AI processors for large language model development, and we believe we can meet this enormous challenge with the current capabilities of our new high-power FOX-XP system with up to 3,500 watts per wafer testing.
We're working on this benchmark as I speak here in the lab right now and expect to complete the evaluation in the next couple of months. Upon successful demonstration of wafer level test results and throughput, we expect they will utilize our new high-power FOX-XP systems for production of their next-generation AI processors starting this fiscal year.
The rapidly growing AI market is still in the early stages, and we see a significant opportunity in this market for our FOX wafer level production systems, as I just discussed. However, in addition, given the unique challenges of testing very high power devices related to AI processors, there's a very real need for a significant amount of engineering qualification and process development, as well as a significant new opportunity for production reliability screening at the packaged part level.
AI semiconductors are amongst the highest power consumption devices in the entire semiconductor industry, with power levels of recent devices up to 1,000 watts or more, well beyond typical processors. These power levels open a new market that requires new unique test solutions.
I'm personally very excited and proud to announce today our acquisition plans for Incal Technology, a manufacturer of highly acclaimed reliability test and burn-in solutions of a wide range of semiconductor devices and markets. They have a particularly strong new product family of ultra-high-power test solutions for AI accelerators, graphics and network processors and high-performance computing processors. Their ultra-high-power packaged part test capabilities, combined with Aehr's industry-leading lineup of wafer level test and reliability solutions, uniquely position us to fully capitalize on the rapidly growing opportunity within the AI semiconductor market as a turnkey provider of reliability and test that spans from engineering to high-volume production.
Incal is in a unique position with intimate knowledge and working relationships with a significant number of AI industry leaders, providing a front row seat to the technology needs of those customers. They are shipping systems today for use by a broad range of companies, with many of these companies projecting needs to move to high-volume production level burn-in of these devices.
Both Aehr and Incal believe there is a tremendous opportunity to grow this business substantially. Incal has world-class system hardware and software architectures and customers that have a high degree of customer loyalty for their products. Aehr brings worldwide sales and support infrastructure, as well as high-volume manufacturing capacity and capabilities that together, we feel, will quickly address customer demand, a very high global growth rate of AI and other high-power semiconductors. We also bring R&D resources, technology and processes and the financial resources to be able to enhance and accelerate new needs that customers may ask for.
This unique combination strongly positions us to capitalize on the significant opportunity within the AI market. Interestingly, we share several subcontract manufacturers and have similar supply chains, as well as our strategy for in-house assembly and final test of our systems. I have known the founders and management team for a very long time, including their CEO, Alberto Salamone, who has been in the test and burn-in business for many years and who will be joining Aehr as an Executive Vice President to lead our packaged part burn-in business.
Incal is located less than 4 miles away from Aehr's headquarters here in Fremont, California with all employees located at that facility. This makes combining the 2 companies simpler and straightforward. We believe that between wafer level and packaged part, the reliability test and burn-in market for AI processors exceeds $100 million annually, and with this combined product portfolio, we have the opportunity to capture a meaningful share of this market within this fiscal year.
So, moving on to the NAND flash memory market, we've been in discussions for several years with multiple flash memory companies related to our FOX wafer level test and burn-in systems. These companies have provided us feedback on the definition and capabilities required for a next-generation wafer level test and burn-in system for their high-volume production road map. This included feedback on our systems, WaferPaks and particularly on our automation using our new fully automated WaferPak aligner. We see the NAND flash market as a key market opportunity for our systems and WaferPaks with long-term potential to also move into DRAM wafer level test and burn-in.
This last quarter, we secured an engagement from one of the major flash memory suppliers to evaluate the FOX-XP system with our proprietary WaferPak full wafer contactors for full -- for wafer level test and burn-in of their flash memory devices. This application is for 100% test and burn-in of devices to be used in high reliability applications such as enterprise storage. This is a benchmark that's going to take us throughout the fiscal year to complete and includes the development of a new high-density WaferPak for production wafer level burn-in of 300-millimeter NAND wafers.
We see this as a multi-year opportunity and expect to have preliminary results and feedback during this fiscal year. Our goal is to come to an agreement for customer-specific development of a test cell with the potential for revenue contribution in our fiscal 2026 that begins next June. We're very excited to have accomplished this critical goal this past year and believe this is the front end of an exciting and potentially enormous opportunity for our solutions.
Another interesting market opportunity is hard disk drive market. One of the new market opportunities for wafer level burn-in is semiconductors used in hard disk drives for data storage. Some of you may recall that in 2019, prior to the COVID-19 epidemic, we announced an order and shipment of our FOX-CP, which is our single wafer test and reliability solution for logic memory and photonics devices. This was a key win with a major customer who purchased the system for wafer level test and burn-in of devices in a very high-volume application for enterprise and data center market. They had forecasted to ramp into production over several years, but the pandemic impacted their plans. After a multi-year product development and qualification process and impact due to COVID-19, this customer that we've now disclosed is a hard disk drive -- in the hard disk drive space, has introduced their product and is now forecasting the production ramp to begin in our current fiscal year, most likely in the second half. We believe this will drive orders for multiple CP production systems and WaferPaks and could even be a 10% customer for us this year.
All right. Turning to the silicon photonics burn-in market, within the silicon photonics market, we shipped the first order from a major silicon photonics customer for new high power configuration of the XP system late in our third fiscal quarter. This new configuration expands our market opportunity by enabling cost-effective volume production test of wafers of next-generation photonic ICs that are targeted for use in the new optical I/O or co-packaged optics market. Nvidia, AMD and Intel are examples of companies that have all discussed the potential for adding optical chip-to-chip communication for performance improvement and power savings for AI processors and high-performance computing chips.
Optical I/O has the potential to be a game changer for semiconductors as it breaks the bottleneck of data transmission bandwidth limitations of electrical I/O. These next-generation silicon photonics-based integrated circuits can require up to 2x to 4x as much power for full wafer test, burn-in and stabilization. Aehr's new high-power system configuration can be used to test and burn-in up to 9 of these new optical I/O device wafers at a time, up to 3,500 watts of power per wafer. This is absolutely unique in the market as we're not aware of any other solution that can test even one of these wafers in a single touchdown, much less 9 of them at a time like we can. While the timing of these devices and volume ramps are not clear, we're watching this market very closely to ensure that we have the products and solutions available to meet the needs of our customers for this potentially significant market application.
Now, let me lastly talk about the GaN market opportunity. This past year, we announced our first order for our FOX wafer level test and burn-in system for gallium nitride, or GaN, devices. While silicon carbide will be the semiconductor material of choice for EV traction inverters, GaN is expected to gain significant penetration in the onboard charging market, as well as other automotive, solar and data center power conversion applications. We're working with several of the GaN market leaders and received a significant number of WaferPak orders throughout the year for gallium nitride reliability test and qualification of our systems. We have now received our first forecast for wafer level production burn-in systems to be delivered during this fiscal year. We continue to be encouraged by this market and believe it will be a significant end market size for semiconductors and has the potential to be a solid market opportunity for Aehr solutions.
Looking ahead, we expect fiscal 2025 to be an exciting year for Aehr. Silicon carbide is poised to be a key contributor to revenue again this year, but we also expect bookings and revenue from across a much broader range of customers and markets, as I discussed. We have a lot of opportunities in front of us, and we look forward to reporting on our progress throughout the fiscal year.
With that, let me turn the call over to Chris before we open up the line for questions.
Thank you, Gayn. Good afternoon, everyone. On today's call, I will summarize our results for fiscal year 2024, as well as the fourth quarter, and then I'll provide our guidance for fiscal year 2025.
Starting with the full year results, we recorded -- reported record revenue of $66.2 million, up 2% year-over-year. Our full year GAAP gross margin was 49.1% compared to 50.4% in the prior year. Our full year non-GAAP net income increased to a record $35.8 million or $1.21 per diluted share, which includes the impact of a tax benefit resulting from the release of the company's full income tax valuation allowance of approximately $20.8 million, compared to non-GAAP net income of $17.3 million or $0.59 per diluted share in fiscal 2023.
In fiscal 2024, we generated $1.8 million in operating cash flows. Our annual bookings in fiscal 2024 were $49 million compared to $78.3 million in the prior fiscal year. The decrease was mainly due to customer push-outs of forecasted orders related to silicon carbide devices due to slower electric vehicle demand in the second half of our fiscal year. Our backlog as of year-end was $7.3 million. With $13.5 million in bookings received in the first 6 weeks of the first quarter of fiscal 2025, we now have an effective backlog of $20.8 million.
Looking at our financial results for the fourth quarter, total revenue was $16.6 million, down 25% from $22.3 million in Q4 last year. WaferPak revenues were $12.4 million and accounted for 75% of total revenue in the fourth quarter, which is significantly higher than the 38% of total revenue in the prior year Q4. WaferPak revenues continue to represent a significant revenue stream for our business due to the strong demand for new WaferPak designs from our existing and new customers as they win new end customer designs and look to meet their market requirements.
GAAP gross margin for the fourth quarter came in at 50.9%, down from 51.5% in Q4 last year. The decrease in gross margin is primarily due to lower revenue, resulting in a higher overhead absorption rate and lower manufacturing efficiencies.
Operating expenses in the fourth quarter were $5.9 million, consistent with $5.8 million in Q4 last year. The slight year-over-year increase is primarily driven by increased headcount-related expenses to support our R&D programs and G&A requirements, which were partially offset by lower professional fees. We continue to invest in R&D to enhance our existing market-leading products and to introduce new products to maintain our competitive advantages and expand our applications and addressable markets.
During fiscal 2024, we have invested significant resources to augment the features and performance of our automated WaferPak aligner and developed a new high-power configuration of our FOX-XP system for volume production wafer level burn-in and stabilization of next-generation silicon photonics integrated circuits.
At the end of the fourth quarter, we released a full income tax valuation allowance and recorded deferred tax assets and a tax benefit of $20.8 million. We released this valuation allowance as we believe it is more likely than not that the company will realize the deferred tax assets.
Non-GAAP net income for the fourth quarter, which includes the impact of the tax benefit mentioned earlier but excludes the impact of stock-based compensation and acquisition-related costs, was $24.7 million or $0.84 per diluted share for the fourth quarter compared to non-GAAP net income of $6.9 million or $0.23 per diluted share in the fourth quarter of fiscal 2023.
Moving to the balance sheet, we finished the year with a strong cash position. Our cash and cash equivalents were $49.2 million at year-end, up slightly from our cash and short-term investments of $47.9 million at the end of Q4 last year. With a solid balance sheet, we'll fund the acquisition of Incal Technology using our cash on hand and common stock. We'll continue to invest in scaling our business and entering new markets and supporting new opportunities.
We generated $1.2 million in operating cash flows during the quarter. We have 0 debt and continue to invest our excess cash in money market funds. Interest income earned during this higher interest rate environment was $592,000 in the fourth quarter compared to $487,000 in the same quarter last year. As of the end of Q4, the remaining amount available under the previously announced $25 million ATM offering was $17.7 million. We did not sell any shares during fiscal 2024. It remains our plan to only sell share against this ATM offering at times and prices that are most advantageous to our shareholders and to the company.
Today, we announced that we have changed the company's fiscal year-end from May 31 to a [ 4-4-5 ] fiscal calendar ending on the Friday closest to May 31. The first fiscal year under the new financial calendar began on June 1, 2024 and will end on May 30, 2025. Our first quarter in fiscal 2025 will end on August 30, 2024. This change is expected to improve the comparability of the company's financial results between periods.
Now, turning to our outlook for the current fiscal 2025. For the fiscal year ending on May 30, 2025, we expect total revenue of at least $70 million, which includes the acquisition of Incal Technology. As we mentioned earlier, we released a full income tax valuation allowance and recorded deferred tax assets in the fourth quarter of fiscal 2024. Beginning in the first quarter of fiscal 2025, we expect to incur income tax expenses. For fiscal 2025, we expect a net profit before taxes of at least 10% of revenue.
Lastly, looking at the Investor Relations calendar, Aehr Test will be meeting with investors virtually at the Needham Fifth Annual Semiconductor & SemiCap Conference on Wednesday, August 21. And then the following week, we'll be meeting with investors in person on Tuesday, August 27, at the Jefferies Technology Summit taking place in Chicago. We hope to meet some of you at these conferences.
This concludes our prepared remarks. We're now ready to take your questions. Operator, please go ahead.
[Operator Instructions] First question comes from Christian Schwab with Craig-Hallum.
Congratulations, guys, on the acquisition. Gayn, as you look at the Incal acquisition, I assume that the company is a growth company. Can you give us an idea of approximately how much revenue of the $70 million is Incal?
Probably the easiest question that I'll try and answer as best I can. So one of the challenges is we need to close it first. And they did about $12 million over the last 12 months, and so, plus or minus a month or 2, even at that same run rate, it's plus or minus $1 million or $2 million or something like that. Candidly, we're taking a pretty cautious stance. But if you use those kind of numbers like $1 million per month for when -- from the time we close it, it's probably a good number. But we -- it gets into a lot of the strategy that's going on. There's been a number of customers that they've been engaged with that have -- we have reason to believe, either directly or indirectly, our engagement is going to help them with respect to their manufacturing plans. And so, we've taken a pretty conservative forecast based upon kind of their current run rates in that $70 million, and we would expect it to grow from there. So, again, just trying to take a pretty conservative stance right now, and we'll know a lot more after we do all the customer visits here over the next several weeks.
No, that's fair. As -- you mentioned on the silicon carbide side that you would expect to qualify this fiscal year with a number of companies in China, given their increased presence in the silicon carbide market and ability to maybe lower prices faster than others and gain further share. How many customers by the end of the fiscal year would you hope to be engaged with?
Okay. Well, I hope to be engaged with fewer customers because I can't -- I feel like we are cup runneth over with the number of engagements because as soon as they become customers, then we stop talking about the engagement and trying to be serious there, actually. I think what you probably meant is how many are we now adding as customers, and if I may put that in your words. I wouldn't put a specific number on it. We've got some internal targets that we're going forward. But I would say several plus, if you look at our current forecast and funnel. There's -- in terms of well-qualified, perfectly capable have fabs, it's well over a dozen, just pure silicon carbide players. These are guys that aren't customers from us yet. It might be close to 2 dozen total if you look at everybody. And we -- as we discussed about it last year, there are probably more questions on China. We are engaged with multiple Chinese suppliers as well and would hope to add one or more of those as customers also within the next 1.5 years or so.
The reason I'm hedging a little bit on the 1.5 years is we know that there are people that are talking about bringing on capacity in second half of '25. And we're just trying to figure out when the first tools would be installed, i.e., I can still win them and maybe install the darn thing in the fall, and that would still be a win for us. But the timing relatively over under our June 1 fiscal year is a bit of a pain right now. So...
Okay. That's fair. And then my last question, on the AI accelerator large language model, can you give us an idea if you're successful there on the new customer? How big could that be?
Yes. We're trying to get our arms around that as well. It's a little weird to talk about who it is or who it isn't, but I've actually just -- and this might get me in trouble someday, but it's not Nvidia. I've been trying to be pretty clear with people because it's just not fair. I guess at some point, if Nvidia ever goes with us now, what am I going to say. But they are a revenue-generating company today. They have customers. They're doing very well. It's pretty exciting. I can't -- there's some discussion about being able to go public with them once we have successfully demonstrated it. They really would have a huge benefit by moving their system-level test burn-in to wafer level. And I can't decide who is more excited about this, if it's us or them, with them cheering us on to please hopefully make this work for them.
This is an interesting one because we are doing some pretty unique things that I'll just share a little bit about, but I'm also holding things to my chest because of competitive reasons. I just don't want to give away any of our secrets. But the idea to actually be putting -- I mentioned 2,000, 3,000 watts on a wafer or 3,500 watts on a wafer. If you're actually close to this technically, you would know that all AI processors at these geometries are -- and lithographies are all 1 volt or thereabout parts. What that means if you're going to do 3,000 watts, you're putting 3,000 amps onto a wafer. People head spin with this idea of putting 1,000 amps, much less [ 2,0000 more amps ] on a wafer. And so what we're doing is quite novel, and we're using the FOX-XP system that we shipped first to optical AI. At the last quarter, we mentioned that, that we were working on something else on the side, stay tuned, but that's what we were alluding to. The development of that system in terms of the power -- being able to put that much power out and remove that much power because you have to remove it all through the wafer itself is totally novel. How we deliver that power for optical I/O is actually interestingly a little higher voltage and lower current, whereas in the AI, it's higher current and lower voltage, but the thermal challenge is the same. So we have a lot of confidence through that. We're working through that with this right now, and we're -- I walked in back just now, was talking to the ops guys. They're working on this, on multiple wafers right now with multiple WaferPaks. So it looks encouraging and stay tuned. I've got my fingers crossed that we can work through all this stuff, and we think we're pretty confident that we can make this work, and the customer is hoping and cheering us on to make it work.
[Operator Instructions] The next question comes from Jed Dorsheimer with William Blair.
Congrats on the acquisition. I guess, just from a framing perspective, is it fair to say that next year is largely going to be driven off of silicon carbide maybe a little bit in gallium nitride and as you invest in some of these other very interesting and high-volume markets?
The way we've actually got the forecast right now is we've taken a pretty conservative stab at the silicon carbide things. And so, I think we're -- we could do well over 30% adding new customers in new markets in that $70 million number. So if you look at the hard disk drive application, I already mentioned, it could be like a 10% customer. The production forecast for the AI is a 10% or certainly at the $70 million. The flash memory will not be. We don't -- I think that's not going to be for revenue. We don't have any in there for this, but we hope to secure an order for maybe the next year revenue. And then, the GaN could -- with the production capacity that we've been showing, that could maybe be a 10% customer or more as well. So you got 10 -- 3 different 10 percenters, and none of those are silicon carbide. And then, I feel like I'm missing one. There's too many of them. But then within the customer base, we are seeing some of the customers candidly we thought we're going to close this year that had moved out in time, but then the fabs are coming. So I think being able to secure that first wave, but several of those guys, at least keeping a single production, if not multiple production systems, could happen by May as well.
So I guess, yes, silicon carbide is still going to be really strong. We think that even within silicon carbide, we'll be more diverse than sort of the -- the 6 customers they have, but only 2 of them really were 10% type customers. We'll actually see more customers that could be material to us, but they are the leading edge to the fabs that will be coming online in '25, late and '26.
Got it. That's helpful. And then, could you just help me connect the dots? If I use -- you said $12 million with 2 months of $1 million, so $10 million for Incal. It would suggest that the core business is kind of -- you're guiding for a conservative or down year at the low end of $60 million. But on the operating income, if it's immediately accretive, which I think was stated in there, are you making a significant investment in the OpEx to cause the EBIT to come down by 6 or so percentage points? Or is that -- is something going on in gross margin?
Yes. You know what, it's -- I'd say it's mostly the prior than the gross margin of it. We've actually made incremental expense investments, some of which candidly was in anticipation of much higher revenue this year, but it was things like the additional infrastructure we put in place in sales, support infrastructure for all the selling that's going on. And eventually, those need to turn into orders as we are now very diversified in terms of the number of engagements at high level. But they obviously need to come to fruition. Otherwise, you put all these dollars in place and they're not helping. So there's explicit direct sales costs associated with that. We also have, in our forecast, it's a little different than last year, the mix of our customers changing with some new customers includes customers that today we're engaged in both directly and with local reps in those countries. And they have a commission structure in them that is higher upfront than later. So we have a pretty material, I think it's $700,000, $800,000 or so in external commissions on what would seem to be the same dollars, but it's actually dollars that are bought by new customers in new markets or new countries that has kind of messed us up a little bit. But good money spent for sure, but that's another one. And then, we've got some of the legal things, legal costs that we've talked about with respect to -- I was going to use their code name with the acquisition, right, that are going on. And there's a few other things with just respect to some profit sharing and some other things are slightly different year-on-year. We do -- we definitely are making investments in R&D this year, both incremental to, I'll call it, the wafer level burn-in product line. We have to get used to thinking about that. But also, we'll be making some incremental investments in the packaged part. We'll talk more about that road map as we close that deal, but some things that we're already contemplating and working on.
It's interesting. We just did our stock planning last week, and we're looking at the R&D programs. And while things like [ Sierra ], the automated liner and some enhancements, the silicon carbide road maps are pretty much in play. We're meeting the customer needs with all the different capabilities that we need. So the bulk of the R&D resources this year are all in the other markets that we've talked about in pure execution against some things, against the GaN guys, the hard disk drive, the flash memory side of things and the AI, all kind of incremental to our platform, so nothing like boil the ocean, but it's kind of fun to watch us being able to start putting more energy behind these other markets.
Got it. And then,, well, that begs, I guess, my final question. So, given these changes, as you start to grow in these other areas, how do you see -- has there been a shift in -- I would assume that once you cover those incremental investments, you'll have a contribution margin that drops. How does that leverage look? Is it at $80 million or $100 million? How do you -- how are you thinking about getting back to that 20% operating margin or above?
Without -- you guys have created your models on there. If you were to think that our OpEx is approximately the same going forward this year. [ So I think ] like it went up, but our revenues didn't go up. I think we put the infrastructure in place majorly to be able to continue to grow, get back to our $100 million plus run rate that we were on without incremental expenses. So, I would be careful ratcheting it down. But if you were to look at similar gross margins and then just incremental revenue, you could draw a line and connect the dots as to when we get back to 25% plus net profits pre-tax.
Okay. The next question comes from Jon Gruber with Gruber & McBaine.
Jon, that's you. He just misspelled your name -- mispronounced your name a little bit.
Yes. A good presentation, a lot of prospects, but what I don't understand is, with the acquisition, all these prospects you get, flash member, 30% in new things, the disk drives. Why is there no revenue growth excluding the acquisition?
Yes. I think the -- you're getting it right. I think [indiscernible] put some numbers together, and that's probably not a bad model to think about. It's really about the push-outs that we saw with respect to silicon carbide ramps that -- things we were expecting people to be coming in pretty strong. And I think we need to write a little nicer than that. And we're just looking at soft forecasts right now. We have multiple customers in our forecast that are going to buy 1 or 2 systems and not a lot of big ones. So our key customers themselves, for example, and again, we've got -- if you just look at the big silicon carbide guys, so let's just back up. So I'm not talking about my customers in general or who they are, but I think if you look at the top 4 silicon carbide customers, they all guided down this year. And so, there have been people that are -- we're wondering how bad it was going to be for us and can we even continue to maintain our growth while they're having a soft year followed by a strong year. So I think we're -- it's the right thing to do right now is to communicate this. If we see strength in the second half come in harder than we are currently conservatively forecasting, then we'll guide up at that time.
The next question comes from Matt Winthrop with [ Equitali ].
Equitable, I don't know. Sort of on a global basis, I have never seen a company turn or you turn as excited 180 degrees from how dour you were the last 2 calls. Is there anything you can put your finger on? Were our cycles shorter? Were you guys super lucky, had a lot of things in the fire that all started to turn? What do you attribute to sort of much more upbeat and such a rapid sort of positive, at least potentially positive outlook going forward?
Geez, Matt, I feel like I'm always a pretty optimistic and upbeat guy. Well, that's kind of weird. And I know you mean that professionally, not on a personal level, but I'll just say on a personal level, I feel like in January and February, we had customers -- when we forecasted last year, we weren't -- our forecast was bigger than what we told you guys, okay? Not that anyone thought $100 million was conservative, but I did. We had these customers that were anticipating those fabs going in. And then by the time we got to November and December, you started to see some of the wheels getting wobbly and people got scared and they just held. We absolutely were -- completed benchmarks or finished with people that we thought would have been buying 2, 3, 4 months ago, and they just pushed out. And it just sort of seemed, as a lot of these things are, if you look at the kind of technology adoption cycle, I can draw it -- but if I can draw it in everyone's mind, you're going up this hill. It actually accelerates a little bit. Then it turns over and goes down, turns around again and then goes up strong. That sort of technology adoption cycle has existed in a lot of different things. And when you turn down that first time, it's pretty scary. And people are like, oh, gosh, that's it. We went from -- 1.5 years ago, our entire business model was built around the crazy idea, one of the things that was driving the silicon carbide, and it wasn't always about silicon carbide, but that 30% of EVs -- or 30% of automobiles in 2030 would be EVs or 30 million. At that time, people were like, come on, is that even possible or not? By around fall, people were saying, oh, it's going to be way higher than that and way sooner than that, which we never repeated. And it was as if everybody and their brother was going to be driving an EV. But we didn't buy into that because we're looking at the fabs, I'm looking at the ground, and there's dirt. And there's -- they're not putting that fab in yet, but people were talking about it with such enthusiasm. And even our customers were starting to get us excited. And then, they went to the opposite like oh, gloom is doom. Everyone is all gloomy. The reality is they're not -- you go and you talk to -- and we have the opportunity, which is the first time in my whole career, I'm talking to my customer's customer, or if you call OEMs, the Tier 1's customer, if you're in the automotive book. We're actually the customer's customer's customer. Sitting with them face-to-face, talking about burn-in times and quality and reliability, and scary, talking about multiple companies' burn-in times and test times and things, not because I'm sharing anything. I can guarantee that. But they're really trying to drive for higher quality and higher burn-in and through the market. And some of those players are going to start building their own silicon, or silicon carbide in this case, to drive for their own quality requirements, very interesting. You see these people and you see their road map, and they're putting in place massive factories, and by the way, way more so outside the U.S. than the U.S. U.S. is really kind of its own thing, and the market penetration is going to end up being less than you'll see in Korea and Europe and certainly China, but even in Japan. But you start talking to the big Japan OEMs as we have, the Korean OEMs as we have, the China OEMs as we have and the European OEMs as we have, you can see through their eyes, this is serious. It's not going to be 30% tomorrow, but EVs are certainly coming. And that gives me reason to believe that we'll be okay and that our business is going to be strong and more and more data to support why you need long way for level burn-in test times. I continue to get reminded of that, including news even in the last few days again, okay?
On top of that, our whole story, if you will, is about semiconductors growing from $600 billion to over $1 trillion by the end of the decade. More and more semiconductors need reliability tests because in reality, they have -- they're not getting more reliable, compound semiconductors. Nobody was using compound semiconductors like they're talking about with silicon carbide, gallium nitride, the elements putting into optical. They all need burn-in. More and more memory, stacked memory, HBM memory, flash memory going to SSDs, processors, AI processors, they all need to be burnt in because they're going into applications where the reliability is not good enough for them to last long term. And as people go to heterogeneous integration or multi-chip modules or whatever you want to call it, it was driving for wafer level burn-in. And now we even see it's beyond that even with the package challenges that's going on with the AI. So yes, if you're reading into that, I'm enthusiastic. We're in a really good spot that's not temporary. This is an upward draft where we saw softness. We're going to be fine. We have highly differentiated sought-after products. And we certainly have the manufacturing capacity inventory to be able to meet those needs. And so yes, we're going to do well.
That, my friend, is a fantastic answer. Keep on plugging, and we'll keep watching. I appreciate everything you do, Gayn.
The next question comes from Tom Diffely with D.A. Davidson.
Gayn, curious, when you look at the book of business you had a year ago, when there's $100 million you thought you would get for the year versus where you are today, I assume most of that was silicon carbide and a lot of that's been pushed out. So I guess the first part of the question is, how far have some of these programs have been delayed or pushed out? Obviously, some of them look like they're about a year behind schedule. And then, the second part of the question is, have they all been pushed out? Or have some of them been canceled?
Yes. So it feels like mostly pushed out. So that's actually -- I like the way you phrased that because it helps me remember. I'm good at remembering what I thought at the time. If I were looking at my forecast and my funnel last year, I had like 3 big guys that were all planning to be buying in the spring, 2 new ones and more from one of our -- the big league guys. I'm sorry, four big guys, okay? One of them struggled to build some products. Another one ended up doing well with their packaged part because of the way the customer mix. The other one had some slowdowns. Another one, the value -- and 2 others -- actually, got you right, there's 5 of them. Two others were in the midst of evaluations, and they didn't end up pulling the trigger because they pushed out their fabs. So I just -- it was $10 million here, $20 million there, $15 million there, and you got to do this. And so, a lot of it just sort of shifted out in time. Every single one of those is still absolutely committed to wafer level burn-in and modules, and their fab capacities have pushed. What they told the Street and themselves 1 year ago is definitely pushed out from that now. But every one of those fabs is -- well, it's not true. Most of those fabs have all been -- continue to be reiterated and re-announced. I think there are some people that might have pushed this fab out a little bit further. So if you look at the OEMs, in some cases, some of the big guys, like if you look at Korea, Japan and Europe, their ramps were always '25-'26. It feels like these early EVs were like foot soldiers, like forward, whatever they call, scouts to test the water, but their big programs are yet to come, and some of those haven't even changed their mind. This is still the exact same schedule they were on, but now it's just getting closer. So it feels like to me that it's about a 1 to 1.5-year push-out of most of those guys. And I believe it will come back. The difference is, I don't think anybody believes -- no one is saying, oh, it's going to be 60% penetration by the end of the decade. They're back more to the 30% kind of number, which is a lot of systems for us.
Yes. Okay. And then the second question would be, think back to a year ago again, and when you think about the car makers themselves, are they all still on the silicon carbide path? Or have some of them decided to stick with silicon a little bit longer?
Yes. So I think it's -- they're more towards silicon carbide than they were a year ago. And I can -- yes, I have specific examples of it. I'd be careful of it. But some of the examples where people were like, well, I'm -- so a lot of cars, as you know, have more than one engine in it, okay? And if you look at 2 years ago, it was very common, people understood that Tesla put the IGBT in the front, and silicon carbide was the first one in the back always. So if you had a single engine, it was silicon carbide. If you had 2, it's silicon carbide in back and IGBT in front. IGBT is silicon for everybody else that's knowing, okay? They had different properties, et cetera. My car that I'm driving, I have a Model S, has 2 silicon carbide engines in it -- engines, the inverters. So we've heard that more and more from the OEMs. They actually preferred to just use silicon carbide. And candidly because costs have come down and availability is up, they can afford to do that. Trying to think.
There's other things I have that's more under NDAs stuff I can't share. But I believe more and more. One thing that shocked me when I was in China is how the China OEM guys really talk to silicon carbide with preference. Now, there's still models that they're going to have a second engine, an IGBT, but it's more and more silicon carbide in all modules as a preference, so I'd say more conviction to silicon carbide and more to modules than a year ago.
Great. All right. I appreciate the perspective.
We have a follow-up coming from Christian Schwab with Craig-Hallum.
I have a quick follow-up, Gayn. We didn't hit China and silicon photonics, whether you expect those to be revenues in fiscal '25 or '26.
Okay. Yes. So I think we have forecasts for China -- well, yes, I guess I just said it. We have forecasts for China in this year. And silicon photonics, I think we have some as well, pretty conservative assumptions right now. Yes, look, I don't think we have it assumed to be at 10% this year. Could it be? Sure. But the problem with the silicon photonics, at least the optical I/O, is -- and again, obviously, there's more than I can share, so job is pretty clean. But those companies that would drive that road map hold those cards close to their chest, right? They're not out -- there's no market. It's -- you asked -- you tell me what Nvidia, AMD, Intel is going to do and the other AI processors, and I'll tell you what the optical I/O market will look like. And they're not talking publicly about it. So we know a little bit more than we can share. We'll just watch. And we'll have to be careful being the canary to let everybody know what's going on. But if people start announcing optical I/O chip to chip, you can just think to yourself, that's good for us. And China right now is all silicon carbide customers. They have GaN too, by the way. I think our current engagements are all silicon carbide today, yes, for China.
By the way, a couple more things on China, a little bit more color for people, right? To us, China is not all one market. And I know people are listening to this, okay? There are companies that are going to build extensions and do things in China that are, say, not Chinese companies, right? And they're very protective of their IP, and they want to be very careful with it. And so, if you sell to them outside of China and they want you to build in China, we love those guys, right? That's not the same, even though it would be in China, okay? Second is that we have companies that are OEMs today that are using our products. Well, they use silicon carbide that's built on our product, and they know it. They drive the test times. They know what's going on, et cetera. They have a high preference for our equipment, and they've talked about potentially dual sourcing in China, okay? Well, in that case, they've said, hey, we want to use your system because we like the same capabilities and all. Well, we like those guys a lot as well. There's other companies that are actually building products themselves. Think of trains and planes or cars. And they want to build silicon carbide. What I can tell you is those companies are so paranoid and so acutely aware of the relationship between burn-in times and quality that they are like, I will dictate a specific burn-in process at a specific quality. It's really important to me. And I like what your system does, and it matters. I like those guys, right? There's other guys that are saying, I want to buy a bunch of systems from you. I've tried some stuff locally. I'm not sure how well it works, blah, blah, blah. I'm like, well, how many systems are you talking about? If you want to buy a bunch from me, I like that a lot. It just makes me a little nervous about the overall IP concerns there. And then, we have companies that say, I'd like to buy an engineering system. Yes, we're not interested. There's a spread. If you're committed to us and you can show us some preference and show us that you're willing to help us protect our IP, even though we have patents all over the place, you get preferential treatment, if you will. So, enough on China. I hope that helps give you a little more clarity, though.
That does. And then, my last question was just a means of potential clarity. You talked about OpEx being flat in aggregate year-over-year because you overspent this year. Does that include Incal? Or is that a comment on your business...
No, my OpEx this year -- I'm sorry, my OpEx this year is higher than last year is what I said. So if you look at the numbers and I think if we just sort of look linear across, you'll find there's maybe $3 million, $3.5 million missing. Like what happened here? We're actually -- that's a result of several areas of expenditures. I forgot to throw Chris under the bus a little bit, too, because we're also spending more money on finance and other things that we did for SOX compliance and stuff. But we have incremental expenses in R&D. We have incremental expenses in legal. We have incremental expenses in commissions. We added more people in sales, and we have a little bit more finance side of things. And then a sprinkling of some bonuses tied around the company kind of to represent those dollars.
Okay. I guess just for clarity, Gayn, for Incal, how much should we assume as their quarterly OpEx?
We haven't done that yet. That's good...
Well, right now, the forecast or outlook, we have already included them in the calculation, in the forecast, in the model right now.
It's a relatively small company. We will -- I mean, I think we've shared about this before, but as our headcount goes up -- they have about 24 employees. They have a lease for next couple of years that's right down the street from us. We haven't talked about synergies. Synergies aren't going to come through people, no way. We need to -- but over time, we don't need that second building potentially, those kind of things, but we're not -- we don't -- we're not needing to scrounge to try and do any expense reductions or things like that, no way. We're going to spend more money with those guys. One last round for folks, Pete. Anyone else with a raised hand?
The next question comes from [ Shahar Cohen with Listed Capital ].
Congrats for the amazing turnaround and the certification into other [indiscernible]. A question about Incal. So first, how much of their current revenue is from the legacy advantage sub-manufacturing, if you can disclose? And B, to what -- the Sonoma family, which as I read in the website, is that the one that's supposed to do the high watt testing? Is that already used in testing of AI application? And is that already used by Nvidia? And did they incur major revenue growth in the last year or so, as one should expect? Or do you expect them to grow significantly in 2025 versus 2024 calendar year?
Okay. All right, Shahar, you've done your homework. So you're going to make me back up and let people know, try and catch up with you a little bit. So Incal is made up of kind of 2 sources of revenue. They have a test and burn-in business, which is made up of really 3 families of burn-in systems, low power, medium power and high power. Their Alpine line of systems is the low power. Tahoe is their mid power. Sonoma is high power. They're all fully compatible from a software perspective. They all have a similar hardware and software architecture, but a very unique platform concept that I think from a tester guy.
The Alpine system uses pin test electronics and power supplies that are shared over multiple burn-in boards and multiple devices on each burn-in board, making it one of the lowest cost, most cost-effective burden systems on the market. We struggled to ever compete with that product line before. Their Tahoe system is a mid-power system, candidly, similar in many of the features to our old ABTS system or our current ABTS system, but it has power supplies and pin electronics that power each burn-in board for more capability and more power with individual temperature control and amazing software.
And Sonoma uses, again, similar pin electronics and power supplies, but per device, allowing them to actually locally generate extremely high currents within millimeters of the device. Very similar to how the application works, which is one of its key differentiators, and that allows them to be able to be used for these really high-power like AI devices.
The Sonoma is, in fact, what has really been growing for them. They have multiple customers on each of their platforms, and there is revenue in all 3 of those segments even within our fiscal year going forward, and we're currently committed to meeting the needs of those customers. We have no plans to abandon any of those road maps, okay? But Sonoma is where a lot of the real growth is, and customers want to pull it into production where we can help them with.
You mentioned something that we haven't talked about publicly, but I'll go ahead and go mention it. They actually also do some repair business for -- kind of as a third-party repair authorization. And we haven't talked about that yet. That's something ahead of us. We've pulled that out of the revenue from last year. So we didn't talk about how much revenue they did in that business, okay? So the kind of $1 million run rate is without that business.
Got you. Really helpful. Any more color you can provide on the Sonoma growth rate maybe that was or maybe expected?
Yes, we're just going to stay [ put ] on this for right now. And I'm not trying to be super elusive. We want to go see all of the customers and be able to build that up, but we'll probably get you more information. In general also, we don't normally forecast too much going forward on all the different product lines, just for competitive reasons, too. But that's the area that I think that we're -- both companies are most excited about to try and help, although there's a bunch of Tahoe customers, too, that are asking for production volumes. So there's -- the mid-power and high-power systems are pretty interesting, and they've been growing, and we would hope that we can help accelerate that growth.
The next question comes from Larry Chlebina with Chlebina Capital.
Again, your AI processor job, once you complete that, could that evolve into possibly getting into heterogeneous PC chips, high-volume PC chips that are all going heterogeneous?
Perhaps, yes. It's very interesting, the dynamics in the test space that have gone on with the advent of system level test, which is highly adapted design for test methodologies, as well as application-specific test methodologies that have really changed the way people look at semiconductor tests. And then, with the heterogeneous integration or the idea you take these chiplets and you take all these devices and you put them all together onto even a silicon substrate, sometimes a silicon substrate has DRAM in it. It actually has active things inside of it. And you have this multi-chip module unlike you've ever seen before, made up of a couple of compute processors, 4 or 6 stacks of high-bandwidth memory, a couple of optical I/O interfaces on it, et cetera. And you think each one of those devices I mentioned today has 100% burn-in. Where are you going to do it? You want to do it at the package level when all those pieces are there? The answer is, not if you can help it. You want to move all that stuff to die level, and wafer is the best way to handle the die. And so, all of this whole topic just gets me excited as a nerdy test guy because that's all good for us. And the -- bringing on -- I keep using their code name, but Incal to be able to help us because they're doing the burn-in of those heterogeneous packages, right? I've seen them. It's pretty cool when you're watching what they're doing. And I can't help but think, well, boy, maybe we can also help some of this stuff go to wafer level. And if not, we got the package. The beauty is I have both.
That's where I was going next. [ Incal ] open up the possibility to go wafer level on some of those projects?
Perhaps. And if not, we got them covered. It's just way better to be able to say whatever you want. But I just think there's opportunities, and you start seeing it all blurring together. You're like, wait a minute, you got this optical I/O wafer level. You got heterogeneous. You've got stacks and stacks of high-bandwidth memory. How can we get back to wafer level? Now, you've got a processor that you can put into test modes and do long-cycle burn-in in a much more scalable, also lower power mode than at the system level. That might not only be an enabler for scale, but might have -- you mightn't be able to even get enough electricity to do it. And you're like, wow, there's just -- it's a target-rich environment. And being able to go and actually sit down with these companies that are building them for their own use or building them for sale or building them for rent or building -- it's exciting.
Okay. One last question. On the flash memory opportunity, I tried getting you on this at the CEO Summit. Is the opportunity in a new fab? Or is it possibly in an existing fab since it requires higher power than maybe existing systems can handle?
This was probably one of the many questions you asked me that I said I can't answer till I answer for everybody.
You said you were going to answer it on the conference call, so here we go.
So I think I can see both. Generally speaking, people think about making big changes when it's time to do a new fab, right? So it's an easier cut to think about the new fabs that are coming online between DRAM and flash over the next 4, 5, 6, 7 years. However, having spent 20 years of my life building memory testers, every 5 years, I was replacing the memory tester I sold them 5 years ago, which sounds crazy, but you get to a point where through parallelism or power or capability, you can't even use the old tool. And this is true with a lot of ATE systems today. Like, I was part of HP, Agilent, Verigy. We were acquired by Advantest. Advantest has the 93000 platform now that we designed in HP back in the late '90s. And today, those 93Ks are fully compatible. People will have -- may have hundreds and hundreds of these on their floor, but they'll buy a new board that goes into that machine each year to meet new capabilities. So in some ways, we built our platform similar like the FOX system. The very first FOX system we built was for flash memory, okay? And people that know our history know that at the time, it was like we were too small and a little too risky, and along came a couple of companies that said, you know what, we're willing to look the other way on your risk because what you have is novel and unique and I need it. One of them happened to be one of the -- or the biggest phone manufacturer in the world for facial recognition, and another one was what now is the biggest silicon photonics company in the world for their platform. That parlayed into multiple different customers, multiple different applications, silicon carbide, now GaN and these other applications where people are interested in using it. But memory is still a core target for us, and we think that we can get into that. Our -- my installed base, those customers have critical technical needs going forward in their road map that is going to require them to make changes. The equipment they have will not work. And at that point, we could displace seemingly perfectly good systems in their fabs with new ones that are better. So I think it can be...
Great if it would be both.
Me too. I'll take one right now. I'll take -- we're excited. This is a big deal to us that this -- people don't make this commitment lightly. And, oh, by the way, there's another one. We're talking about expenses. I'm going to spend some money on this flash development this year. It's expensive, what we're going to be doing. And we're going to build up some WaferPaks. We're going to be doing some technology. We will do some prototyping. We're going to put a bunch of manpower on it. We may not make a dime this year, and it will be -- all that OpEx is money worth spend because this -- I think this market needs it. I think flash first than dram, and I think we're in a great position architecturally to be able to address that. And so, any one of those deals is enormous.
Well, it's worthwhile doing, that's for sure.
There are no further questions in queue.
All right. Folks, thank you very much. I know that we ran a little longer than normal. We really appreciate everyone's time. We'll figure out how to make these as concise as possible as our story is no longer focused on a narrow market or 2 and a couple of few customers. So we'll find our way to be able to summarize and make it easier to digest. But we're really excited about this, excited to head into the new year. And we welcome our new friends from Incal. We're throwing a lunch for them in a couple of days, and we're excited to host them to come over and meet the crew. And we'll keep you guys updated on a quarterly basis.
Have a good one and take care.
This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.