ACM Research Inc (ACMR) 2020 Q4 法說會逐字稿

Good day, ladies and gentlemen. Thank you for standing by, and welcome to the ACM Research Fourth Quarter and Fiscal Year 2020 Earnings Conference Call. (Operator Instructions) As a reminder, we are recording this call. If you have any objections, you may disconnect at this time.

Now I will turn the call over to Mr. Gary Dvorchak, Managing Director of The Blueshirt Group. Mr. Dvorchak, Please go ahead.

Good morning, everyone. Thank you for joining us on today's call to discuss fourth quarter 2020 results. We released results after the U.S. market close yesterday. The release is available on our website as well as from newswire services. There's also a supplemental slide deck posted to the investor portion of our website that we will reference during our prepared remarks.

On the call with me today are our CEO, Dr. Dave Wang; our CFO, Mark McKechnie; and Lisa Fang, the CFO of our operating subsidiary, ACM Shanghai.

Before we continue, please turn to Slide 2. Let me remind you that the remarks made during this call may include predictions, estimates or other information that might be considered forward-looking. These forward-looking statements represent ACM's current judgment for the future. However, they are subject to risks and uncertainties that could cause actual results to differ materially. Those risks are described under risk factors and elsewhere in ACM's filings with the Securities and Exchange Commission. Please do not place undue reliance on these forward-looking statements, which reflect ACM's opinions only as of the date of this call. ACM is not obliged to update you on any revisions to these forward-looking statements.

Certain of the financial results we provide on this call will be on a non-GAAP basis, which excludes stock-based compensation, a loss relating to the change in fair value of a financial liability and an unrealized gain in trading securities. For our GAAP results and reconciliations between GAAP and non GAAP amounts, you should refer to our earnings release, which is posted on the IR section of our website.

With that, let me now turn the call over to David Wang, who will begin with Slide 3. David?

Thanks, Gary. Good morning, and welcome, everyone, to today's call. 2020 was a very productive year for ACM. We have faced several challenges, such as the COVID-19 pandemic and trade tensions. We overcame these challenges through hard work and good execution and a positive tailwind in China semiconductor industry. We focused on what we could control. We managed our supply chain. We executed our customer deliveries. We expanded capacity. We introduced new products, and we made great progress with new customers.

We also moved forward with our long-term facility plan in Lingang and completed a deposit for employee housing development.

Before getting into detail, I would like to thank our employees for their hard work and dedications. I also want to thank our customers, partners and shareholders for their continued support and confidence in ACM Research.

For the full year, we had a record revenue of $156.6 million, up 46% and a record shipment of 182 million, up 58%. Non-GAAP operating margin was 17.3%, and we ended the year with a $71.7 million in cash and additional $28.2 million from our holdings of SMS and STAR Market shares. Our success starts with our customer.

Please turn to Slide 4. We have 5 major front-end customers across CD-ROM, foundry and DRAM as well as several back-end wafer packaging and assembly customers. Our new customer, manufacturer call and analog devices, they choose ACM technology to optimize their own production capability. We'll help them achieve the operational excellence they needed to compete in today's global semiconductor markets.

Let me discuss our key customers. I will start with the Shanghai Huali and Huazhong Semiconductor, also known as Huali Huazhong Group. They are leading advanced foundry in China and our core customers in 2020 had 37% of total sales. They are great customers and neighbors with production near our Shanghai headquarters. Shanghai Huali is a leader of a multiyear capacity expansion, making good progress in 20 millimeter. While our semiconductor is adding capacity and is truly edge fab in Asia, their major products are CIS and power devices, and capacity will be more than 80,000 per month by end of this year, up from 30,000 per month at the end of last year. We are participating in both projects. We have several first group -- first demo system in Huali for some of our newer product offerings.

Next, YMTC. YMTC is our seventh largest customer, contributing to 27% of the total sales in 2020. We are working closely to support YMTC 3D mass production line in Wuhan. Public reports say that YMTC give its production capacity to 50,000 wafer per month in 2020, and they are expected to add another 25,000 to 50,000 in 2021 and convert from 64 layers to 128 layers. We believe they are, YMTC, the largest single-wafer Western supplier. ACM too are being employed in a significant number of cleaning fab. So we continue to expect strong demand from YMTC as they scale their production capacity in Wuhan and other major facilities to come.

ACM's further largest customer was SMIC, the largest foundry in China. It is 12% of the sales in 2020, up from single digits in 2019. While it's in Shanghai and the flexibility to deliver SMIC, our business level also depends on other new suppliers to gather license to ship their own tools to AMC. We are well positioned with a range of the tool SMIC, but are factoring in just a small amount from SMIC in our 2021 outlook.

Now let's touch on our other significant customers. I will begin with SK Hynix, the #2 global DRAM supplier. SK Hynix was ACM's first major customer, a strong testament to our proven fast cleaning technology and ability to improve production yields. We started with just a few production steps at SK Hynix. Now our tools are used in more than 20 steps as a progress to more advanced nodes. SK Hynix was less than 10% of our sales in 2020, down from about 20% in 2019, as expected due to pause in DRAM cycle. SK Hynix implemented several technology updates in 2020, and we expect them to return as a 10% customer for 2021 and beyond through technology upgrades and capacity additions.

Next, XMP, a non-China-based entrant to the DRAM industry. We deliver our first SAPS-V tool to them in later 2019 and recommend revenue in Q4 of 2020 upon acceptance. They select us for our compelling value preparation, proven ability to improve yields and a solid technology road map. XMP is in the early stage of a multiyear production plan. We are positioned to participate in the SAPS, Tahoe, very critical cleaning pools, ECP and other tools when they scale up in 2021 and beyond.

Finally, we have 2 new analog power IC customers. We delivered multiple first tool in the third quarter -- third and fourth quarter of 2020 and recognized some revenue in Q4. The first tool, including our semi critical tools, scrubber, thin wafer, backside cleaning tools and ultra wet bench SAPS cleaning tools and Ultra FM Furnace tools for (inaudible) cleaning process. With this, we have penetrated 2 of the 5 key cleaning edge nodes and other PIC and CIS manufacturing in China. I'm happy to see we are very active with all of the remaining 3 players and expect to receive orders from them this year.

We believe these customers alone represent a significant opportunity for ACM. Many of them are still in early or middle stage of multiyear capacity expansion and are buying just a friction (sic) [fraction] of our products. We think our current customer base alone can support solid growth for years to come, as this customer add capacity with our tool in additional production fabs and buy more ACM products. Innovation will enable us to capture new avenue of a large market opportunity.

Please turn to Slide 5. As an emerging supplier, we're focused on delivering advanced technology at competitive cost. We have a strong R&D team in Shanghai and Korea, and one of the largest service teams for semicon equipment and employees in China. This puts a critical mass of near some of the newer and the larger semiconductor projects on the planet. We consider this a great opportunity. We have not yet stretched the scope the next few years, a landgrab period, to expand our product line and scale our business as new major semiconductor products build their -- produce or build their business in China. We intend to gain market share by expanding our product line and gaining new customers both in China and around the world.

We estimate our current product portfolio address about USD 5 billion of the global wafer clean market. This includes DRAM, 3D NAND, foundry, power devices and analog applications. Our cleaning tools adjust about $2.5 billion or 80% of a $3 billion total wafer cleaning market. We start with our flagship, SAPS, TEBO and Tahoe, of course, coupled with our semi-critical cleaning tools.

Our new product add another $2.5 billion, including $1.7 billion by Furnace, $5 million by our ECP and more than $30 million by our strategy polishing products, wafer manufacturing and other advanced packaging process equipment, including coder, developer, wet bench, scrubber and [working] receiver.

In 2020, we spend 12% of sales or about $18 million in R&D, up 55% from 2019 levels. That spending enable us to expand our offering beyond SAPS and TEBO and Tahoe, semi-critical tools, ECP, and most recently, Furnace products. Our ECP tools are off to a very good start. In 2020, we delivered a handful of first tool to front-end and back-end packaging customers and had a greater revenue contribution for the year.

Our key products, including ECP AP for advanced packaging and ECP MAP for advancing copper interconnections, ECP TSV for (inaudible) for front end. Our secret sauce is the ability to provide uniform plating on outer center layer. For advanced packaging, we have developed a proprietary high-speed carbo plating technology with the patent pending to significantly improve copper pillar plating performance. To place metal film with deep (inaudible) or tools, SAPS (inaudible) 20 micron at a high speed and better reformatting is a major challenge around 3D plating applications today.

Historically copper plating for pillar at a high trading rate due to the mass transfer limitation that reduced preparation rates and generally the uneven top profile of a pillar. Our new high-speed plating technology solved the mass transfer challenges while achieving a faster pillar for profile and delivering a faster height in quality less than 3% as well as the high throughput. The CCGT is high-speed plating rate, together with our SAP copper polishing pool, will become major wining products in coming 3D advanced packages.

We are also excited about our Furnace opportunities in 2020. We delivered 2 foster tools to 2 companies, and they plan to deliver several more in the first quarter this year. Furnace has been jointly developed by our Korean and Shanghai team. Both teams have worked seamlessly to solve their alpha and beta tool issues and the teaming process to meet customer requirements.

We have made a significant progress in improving this process, including HDL, cleaning and [answering] and a high vacuum palette of cleaning process. Our teams continue to develop at (inaudible) and high-temperature oxidation process. Our next step is develop such autonomous layer separation, ALD process, which is the most challenging and promising product for the advanced manufacturing nodes. We have successfully implemented our multiple product strategy and expanding our product portfolio from wet products as wet cleaning SAP and SAT, with joint product and vertical Furnace, which make our first entry to dry products and more to come in the near future.

For 2021, to prepare for new opportunity ahead, we plan to accelerate our R&D to about 14% of our sales, redoubling total market addressed by our products from $5 billion today to more than $10 billion. We are too early to offer details. I can see that our team has begun work on serving new major products that can get us close to the $10 billion addressable market growth in the next couple of years.

Next, I will comment on our efforts on winning major first tier new customers. Our investments in our global sell team are making steady progress, we hire a US-based team of sales in April 2020 as well as several other key hires. During the year, we delivered our very first win in a few labs, with a lab of the equipment player that needs to separate cleaning with its own process. Thanks to the greater execution of service team, the customer formally accept the tool, and we book the revenues in Q4.

The team is actively building a sales pipeline with regular engagement, such as advanced technology discussion in the U.S. and the Taiwan players. We are confident that, in fact, TEBO, Tahoe and high-speed copper plating SAT as well as SAPS, we can secure wins at one or more major first-tier of customers during 2021.

To fulfill our growth ambitions, we will need capacity. Let me update on our progress. Please turn to Slide 6. Our original facility in Shanghai remains; the headquarter is in Shanghai. It includes our R&D, SG&A and prototyping and production of new products. We stated -- we started production at our second factory in Chuansha during September 2018. We quickly finished and added a second floor to production in Q3 2020. We now have 300 million of annual revenue capacity at Chuansha. We believe that we can further expand to more than 5 million of revenue capacity by leasing additional space.

Our new Lingang facility will include production and R&D with a planned 1 million square space of floor space. We will have ability to increase our production capacity to 1.5 billion. During 2020, we purchased 50-year land rights in Lingang. We expect additional architectural and design work in the first half of 2021 with initial production expected in later 2027 (sic) [2022].

Ventures in our Lingang operation is low-cost employee housing. While not common in the U.S., it is expected as a practice for large companies in China, particular in Shanghai. We offer housing to attract and maintain key and experienced employees. We train many of our key engineers and managers for more than a decade. We need to keep these key employees to secure our future growth and technology development. During 2020, we deposit $40 million towards the purchasing of 152 housing units. We paid for -- these are over half of the amount from ACM Shanghai cash balance. We financed the other half with a 10-year loan facility.

Before I provide our 2021 outlook, let's discuss the status of the Star Market IPO of ACM Shanghai. Progress is steady but slower than expected. Soon after '18, Shanghai complication of 2 -- the Shanghai Stock Exchange Commission or SSEC from September 30, 2020, events in the U.S. caused a delay in IPO. It is stated with the short sale report issued on October 8, 2020, followed by a class-action lawsuit by December 21, 2020. We have previously described our disagreements with the short seller report, and we, therefore, believe there is no substance to losses, which is largely based on the short seller report.

The SSEC, which is not familiar with this kind of report and a lawsuit, want to better understanding the U.S. legal process. As we have also previously described, we responded with a 5 points verification report to the short seller report. And that [venue] that issued the report was accepted by SEC on December 11, 2020 and released publicly in both China and the U.S. Before our administration application should be filed with the China Securities Regulatory Commission, or CSRC, however, the civil suit was fined, and we ran that as SSEC would again require additional information.

Our team is now in the process of providing secondary response to report to SSEC to plan the class action lawsuit. We are confident that we will go over this hurdle and complete our IPO. But SSEC does not provide us with a timetable for this review that would enable us to predict the precise timing.

Now let's move to our 2021 outlook on Slide 7. Our values reflect optimism about our growth prospects for 2021. The report that we offer preliminary guidance in early January, which is unchanged. We expect revenue to be in the range of $205 million to $230 million, which present 39% annual growth in [cleaning] business.

Our outlook 2021 is based on several key assumptions. First, the global COVID-19 situation improves in the coming months; second, U.S.-China trade policy stabilize; third, a range of spending scenarios for the production lending of key customers; fourth, variance in the trajectory of DRAM recovery; and finally, a range of outcomes for the timing of customer acceptance of the first tool.

Our results and outlook demonstrate the successful execution of our strategy. Our strong growth is providing the capability to accelerate our R&D spending in new products. We are building a global sales and marketing resource to penetrate new customers in new regions and we're scaling production capacity to support our long-term growth plan. We are increasing operation spending to take advantage of their growth opportunity ahead. Our mission remains to become a major equipment supplier to the global semiconductor industry.

I will now turn the call over to Mark to discuss the financial results in more detail.

Thank you, David. Good day, everyone. Q4 was another strong quarter, capping off a tremendous year for ACM. For the full year, we grew our business significantly with increased market share, good customer exposure and expanded production at our second factory. We launched several major product lines, including the Furnace, semi-critical cleaning, introduced a number of extensions for SAPS, ECP and advanced packaging. We closed the year with a solid balance sheet, including $72 million in cash and an additional $28.2 million of trading securities of our SMIC stockholders.

Now I'll put some detail around our full year 2020 results. Unless I note otherwise, I will refer to non-GAAP financial measures, which excludes stock-based compensation, change in fair value of financial liability and unrealized gain in trading securities. A reconciliation of these non-GAAP measures to comparable GAAP measures is included in our earnings release.

First, highlight to the results for the full year 2020, shown on Slide 8. Revenue was $156.6 million, up 46%. We break out 10% customers annually and provide occasional color on our quarterly calls. As David noted on Slide 4, we had 3 10% revenue customers for 2020. First, the quality Huazhong Group together accounted for 37% of revenue versus 26.5% in 2019. That grew by 103%, a big contributor in growth as we expanded penetration from Huali Shanghai fabs to additional factories of Huazhong and Wuxi.

YMTC was our second largest customer at 27% of revenue versus 27.5% in 2019. Our revenue from YMTC increased by 42% year-over-year. To help expand the YMTC to expand their wafer starts to $50,000 per month and summit share gain into additional cleaning steps drove the growth.

SMIC was 12% of revenue versus single digits in 2019 and the participation with SMIC improved nicely in 2020, due in part to successful sales effort and SMIC's recognition of our ability to scale.

SK Hynix was a little less than 10% of revenue versus 19.8% in 2019. This was as expected due to a pause in DRAM spending.

Revenue from wet cleaning and other front-end processing was 87% of 2020 sales versus 84.6% in 2019. Revenue from advanced packaging, other back-end processing tool services and spares was 13% of sales in 2020 to 15.4% in 2019. Total shipments were $182 million versus $115 million in 2019.

Gross margin for the year was 44.5% compared to 47.3%. This was within our normal expectation of 40% to 45%. We expect gross margin to continue to vary on a quarterly basis due to product mix and manufacturing utilization.

Operating expenses were $42.7 million compared to $29.5 million. The increase was due to higher R&D spending on new products, increased sales in North America and strained IPO-related costs. Operating income was $27.1 million compared to $21.4 million; operating margin was 17.3% versus 19.9% a year ago.

Non-GAAP results exclude stock-based compensation. Two additional line items below the operating line. The first is a change in fair value of financial liability as a nonoperating noncash book loss of $12 million. As described in previous calls, this was related to investments in ACM prior to our 2017 IPO and restructured in Q3 of 2020 in connection with the STAR Market IPO. The liability was terminated upon our issuance of an equity warrant, after which it became a balance sheet item and no longer impacted our income statement in Q4 and beyond.

Second is an unrealized gain of traded securities of $12.6 million from our investment in SMIC prior to the STAR Market IPO. The investment was mark-to-market at year-end, and the gain reflects the increase in value from the original IPO price. We will exclude this item from non-GAAP results until the gain is realized if and when we sell the shares.

Net income attributable to ACM Research was $23.8 million versus $22.5 million. Net income per diluted share was $1.12 compared to $1.17 in 2019. Tax items and the effects of foreign exchange fluctuation on operating results provided a net benefit of $0.9 million and $4.7 million in 2020 and 2019, respectively, or $0.04 per share and $0.25 per share, respectively.

Now for the fourth quarter, as shown on Slide 9. Revenue is $45.6 million, up 85.2%. Total shipments were $67 million versus $25 million in the fourth quarter of 2019 and $59 million in the third quarter of 2020.

Gross margin was 43.3% versus 50.7%. Operating expenses were $13 million versus $8 million. Operating income was $6.7 million, up 39.6% from $4.5 million. Operating margin was 14.8% versus 18.3%. Unrealized gain on trading securities, excluded in non-GAAP for reasons mentioned above, was $3.6 million in the fourth quarter of 2020.

Net income attributable to ACM Research was $6.2 million versus $4.6 million in 2019. Net income per diluted share was $0.29 versus $0.23 in 2019. Tax items and effective foreign exchange fluctuation on our operating results provided a net benefit of $0.9 million or $0.04 per share in the fourth quarter of 2020 versus a net benefit of $1.1 million or $0.05 per share in the fourth quarter of 2019.

Now I will review the balance sheet. Cash balance was $71.8 million at the end of 2020. Short-term borrowings at year-end were $27.7 million versus $13.8 million in 2019, and long-term borrowings were $19.7 million.

Total inventory was $88.6 million at year-end versus $44.8 million in 2019. The increase in inventory was driven by a couple of factors: First, finished good inventory. For ACM, that represents first tools delivered to our customers for evaluation and awaiting for acceptance. Finished goods grew to $32.4 million at year-end, up from $19.3 million at the end of 2019. The remaining inventory growth was mainly due to work in process and raw materials to support our sales growth in 2021.

Cash flow used by operations were $13.5 million for the year. Cash was used to fund our growth in 2021 and to deploy first tools and new products for key customers. Capital expenditures for the year were $5.5 million versus $2 million in 2019.

In 2020, we spent a total of $50 million on Lingang-related investments. This includes the $9.7 million for the 50-year land rights and $40.2 million in deposits for employee housing. For 2021, our base case for capital spending is $10 million to $15 million. Our 2021 investments will be balanced between capacity increases at our second factory in Chuansha, investments to support our R&D programs as well as planning and some initial spending on Lingang. We expect to adjust our capital spending upwards after our China IPO.

To conclude, we continue to execute on our strategy. We are participating in the growth of major new IC fabs. We are ramping production and continue to develop and deliver innovative products. We're positive on our opportunities in China and our expansion outside of China.

Let's now open the call to any questions that you may have. Operator, please go ahead.

(Operator Instructions) We have first question from the line of Patrick J. Ho from Stifel.

Congrats on a nice finish to the year. Maybe, David or Mark, looking ahead at local Chinese domestic spending, we're looking at potentially another strong, robust year. Can you just give a little bit of color of the breakdown between your thoughts on memory spending versus the trailing edge foundry logic spending and whether you see potential growth in both areas? Or is it going to be weighted more on one segment or the other?

Okay. Thanks, Patrick. Actually, we have very good availability in our basic 3D NAND and also [1Q3] and also Huazhong Huali. And so we see that continuing to -- quickly spending their fab, both in the advanced 3D NAND, also their memory side, also in the logic funding side for Huazhong Huali group, right? And plus, they also acquire their -- so this 5, what we call the second-tier customer. And there, we got 2 already. And this year, we have companies who will get rest of the 3 additional second tier, I call it, training edge, the customer tier. So we're building favor collectively. And I still say probably -- more than 50%, 50%, they come from our top or first 3 customers. And we're seeing that there are 5 additional new customers expanding quickly. We cannot give a concise number of the next -- this year, what the region will be. But obviously, of course, they're going to expanding their fab.

Mark, anything you want to add on?

Yes. No. Thanks, Patrick. I mean I think you asked about the relative growth versus memory and trailing edge. And as you know, YMTC is a 3D NAND big customer, so we're expecting to grow it there in their capacity adds. David mentioned Huali Huazhong. SK Hynix, David, in the prepared remarks. We're anticipating a DRAM recovery from them. And so they were less than 10% in 2020. And so we'd expect them to kind of get -- indeed, get to show some decent growth here in 2021.

Great. That's helpful. And maybe as my follow-up question for you, Mark. As you get the new facility and factory ramped up for production, how do you look at gross margins given that there are typically startup costs, some underutilization to begin with? How quickly do you get that within to your normal, I guess, corporate ranges as that facility ramps?

Yes, you bet. So Patrick, we're not going to modify our gross margin outlook, kind of the 40% to 45% range, right? There's a lot -- obviously dependent on the product mix, and that's really the biggest driver. Relative to production capacity, we said a bit here. Chuansha, we're adding capacity pretty quickly. We've got really pretty good demand throughout the year. The Lingang facility, that -- we're looking at that for the end of 2022 is when that would start kicking up. So we don't really anticipate any impact to our gross margin from that Lingang. That would be out there in 2023 and beyond.

We have the next question from the line of Suji Desilva from ROTH Capital, please.

David, Mark, congratulations on a strong end to the year. A quick question on the STAR listing. Can you just give us a sense qualitatively the level of detail the SSEC requested in this additional review versus the original report you filed? Just give us sort of some sense of magnitude here of how much more they're looking for.

Okay. As I just stated, right, this short seller report had come out, and we spent almost 2 months to finish the report. And then we got accepted, and almost we're starting to acquire we call it registration for CSRC, but then there's a lawsuit that come up. And anyway, now we're taking care, also we hired their very experienced lawyer in the U.S. And our lawyer told us pretty much no substance in this lawsuit right now. And so we're going with -- again, in response to the report to SSEC to give them the detail what is going on, what is legal process in the U.S. I think it probably is going to file this report very soon, and then we're waiting for their response, how they think about this. If it's a positive report, then we can start continuing the registration process in the CSRC.

So that's the timing right now. But again, we don't know how long for there to resume second report. And that's why we couldn't give precisely timing of the IPO. But I think we're confident. I think we get over almost 90%, 95% down already, and the remaining 5% and we are pretty confident. We should be a success with IPO based in China.

Okay. And then perhaps on the financials, the tools you have placed with customers entering '21. Maybe -- I'm curious, do you have a higher percent of visibility from those tools into your guidance for '21 versus, say, a year ago, given that you have more customers and more products?

Yes. I can see that this year obviously is better than last year, right, and looking at our projection. And also, as you are aware, we're very busy on the Q1, Q2. Even we're really stretching our kind of ring plates and very [frame of] the manufacturing capability. So we start to hire more people and also training more people and to final path also to the start of the year. So at this moment, it's very busy Q1, Q2 including Q3. So we're very -- I called it confident in -- for our projection for this year.

Yes. And Suji, you could obviously look at the level of the finished goods inventory, right, at the end of last year, the end of 2020. And so it grew from, what, $19 million to about $32 million. So -- and that's carried at cost. So assume our corporate gross margin for that to get an idea for how much revenue visibility would be there.

We have our next question from the line of Charlie Chan for Morgan Stanley.

So my question is about your China customers' contribution. You mentioned about YMTC and Huazhong. But how about SMIC? And because the company also announced they want to spend around USD 4.7 billion CapEx. Do you think that is going to be a significant growth driver for you for this year? And what do you think about their capacity expansion progress given the SAPS issue?

Okay. I think we're pretty well positioned with SMIC, and we have worked with them for the last 3 or 5 years. And we do have a product and so then in cleaning and there is our -- I got key process in SAPS. And also, we also have, I call it, semi-critical process tool also being sold to SMIC. We also do have copper plating and also our TSV tool, right? And plus also feature -- also position our Furnace tool into the SMIC too.

So I think we're very positioned for the market now. And look in their announcement, they're expanding 10,000 per month of their, call it, revenue expansion for their probably 128-nano technology. So we are positioned for that, too. Again, if this licensing can be solving quickly, we can see quite a tremendous opportunity in front of us. However, this moment, we're very cautious watching and [manipulating] the stuff. And obviously, without their license went down and for them to be expanded quickly, it's kind of difficult, right? So anyway, this moment, we just help and are working closely and -- even just our customers. So we're making good, supporting service for the company.

Yes. Charlie, one thing I'd add, right? I mean it's obviously a bit of a wild card, right? I mean we've got -- David mentioned really good visibility with a number of our top customers. But for SMIC, we're not factoring a significant amount in our outlook. It really depends on their ability to get tools from other U.S. manufacturers. So -- and that demand in China is pretty strong across the board.

And we also think that it's mixed depending on how much capacity they add and some of the other manufacturers may add, may step up their spending. It's hard to tell, but for them, we're not going to factor too much into our outlook.

The next question comes from the line of Quinn Bolton from Needham & Company.

So let me -- my congratulations to the strong finish and the nice outlook for 2021. David, wanted to just kind of follow-up on Suji's question around the SSEC review. I understand you're about to file your response with more information about the class action lawsuit. And you don't know how long it will take the SSEC to review that. But once you receive that -- or once the SSEC has completed its review, can you walk us through how long does it typically take to get through the CSRC registration process? So folks can have a sense of time, that once that SSEC review is complete, what the time line for an IPO might look like? Is it order of weeks? Is it a couple of months? Any thoughts you could provide would be helpful.

Okay. Quinn, very good question. And actually, as I said, when the SSEC finished a review for our second report, then we're getting into the so-called registration for CSRC. And by their -- I should say, so far, their statistic data show averaging registration time in CSRC, about 45 days to 60 days. So that will be their registration time period. And after that, we'll probably spend the other decision down because another 2 weeks or 3 weeks, the road shows that you can -- kind about the 2 to 3 months, right, IPO time. That's after we finish the acceptance or approving assets we're seeing for secondary report. Did I answer question?

Got it. It's perfect, yes. And the second question is -- a very nice job on the shipments in the fourth quarter. Mark, I believe I heard you say $67 million. Obviously, that gives you some pretty good visibility into 2021. I'm wondering if you might be able to break down for us how much of that $67 million were your kind of traditional critical clean tools versus some of your newer semi-critical Furnace, ECP tools. Is the majority still the critical clean? Or are you starting to see a much broader kind of composition of the shipments you're making?

Yes. Maybe I couldn't give you the sort of detailed numbers. I can give you general. Probably cleaning tools, they occupy about 85% -- 80%, 85%, I call the total cleaning is our, I call, flagship, also semicritical together. The rest of their about 15%, 20%, is our new products coming, copper plating. And also, we have application in the packaging application. So that probably gives you the ballpark, 80-20 divided.

The next question comes from the line of Chi Tsai from Jefferies.

So you reported to deliver $188 -- $182 million shipments in 2020, and I think that would indicate $31 million backlog in 2020. And I think if I'm getting right, I think we now have more than $60 million backlog. And do you see this backlog, well, expanding in 2021? Or you are expecting the similar range?

Yes. Yes, let me just correct a few things. So the way you would look at the value of our tools that are awaiting acceptance, right? You'd want to look at the finished goods inventory level, not the number that we shipped in the year. But that finished goods inventory was a little over $32 million, right? So that's carried at cost. So you would assume the corporate gross margin to get an idea for the level of potential revenue that's in that.

And so I think what you may be asking, for 2021, we don't really give guidance on our overall shipment. But we would anticipate strong shipments again -- shipments to additional delivery of demo tools this year.

Yes. Well, obviously, this year, there's more new customers coming in, especially, we have a new product in SAPS, and also our Furnace product and also additional new customer come in, we're expecting more of their, I call, deferred revenue will come out this year, too, which is recorded -- not recognized, but immediate upon shipment, but we recognize revenue upon acceptance. So this year, I expect the end of this year, much more of the first tool to be deferred than the next year.

I see. My second question is that, can you kind of bridge the outlook of 2021 revenue in first half and second half? Because I think a lot of capacity expansion plan, like additional expansion plan, were made in the second half last year due to capacity shortage. So I think more capacity will be delivered in second half. So I think, can you give us some idea on how first half and second half will look in this year?

Yes. I think I can address that. I mean we typically just give a full year outlook. But Q1 is typically our seasonally weakest quarter, right? You got Chinese New Years in there. So we get a nice -- generally a good snapback in Q2. Q3 has traditionally been our biggest quarter, and then a slight decline in Q4. So that's how we're looking at it. And I think you're correct, a little bit more in the second half than the first half, kind of balanced that way. But nothing traditionally out of the ordinary from the seasonality you've seen in the last year.

The next question comes from the line of Mark Miller from Benchmark Company.

You indicated you expect the return of Hynix to the 10% customer. Is that more of a second half or a first half type return?

Okay. I think that -- I think, look, probably in the second half of this year. And we do have some order coming in, in Q1. And also look at the market, and the people project that DRAM will be quite recovered compared to last year. So we're expecting some other order coming into the second half of this year.

Your sales and marketing expenses were up significantly -- I'm sorry, your sales expenses were up significantly, sequentially as well as R&D despite somewhat lower sales in the fourth quarter. Do you expect this trend to continue into 2021?

Yes. I think we're using the number of our sales marketing, about 9% to 10%. And that range will keep -- continue. However, this year, we do add additional, probably, 3 points and more for R&D. So it totaled about 14% and plus R&D efforts because as we mentioned, it's a very pivotal year. So it's a big year for ACM flow. And it's critical that we are positioned strongly in the Asian market, and that's why we're investing more to develop new product come out. And those give more spending for R&D.

So you said sales will be about 10% roughly?

Yes. It's 11%, 10% sales marketing. And then we're funding more of the R&D business until the G&A. That's the total spending for the brand.

Yes, Mark, just to make sure you're aligned, this is the non-GAAP, right? So we're excluding the stock-based compensation, those percentages that Dave was giving you.

Yes. You mentioned stock-based. What was stock base comp last quarter?

Q4 stock-based comp was $1.3 million, $5.6 million for the year.

We have a follow-up from Charlie Chan from Morgan Stanley.

Yes. Charlie, I think we cut you off somehow or maybe we lost you.

Yes, yes. Sorry, I got disconnected, working from home. So maybe another follow-up is for you about your international customer progress. I know Hynix probably will be dated this year. But how about other major Chinese -- Huazhong, TSMC or other U.S. customers? Any progress here?

Great. Charlie. So let me maybe elaborate more, Charlie, on our next big plan, coming to the outlook for the customers. And we're always looking 3 and 5 years ahead, what's the technology trend today in the industry. People talk about all the gain all along and also the 3D probably come to the real limitation to nano, when nano is really hard and also very expensive, too. So people move to 3D. Now we can say 3D NAND is one of the structure people thinking. People are talking about 3D DRAM and advanced packaging.

So I look at the trend ahead. ACM is really well positioned for this huge technology achievement. As a part of our cleaning tool, our SAPS, TEBO is really good for cleaning small particle. TEBO is only technology today in the world that can reassess tiny [future] -- or tiny particle inside deep change or in the [rear]. That's only available as to the market, and you can clean those deep or tiny corner particles. And by other technology like in [Jet3], you never can get out. And also, the [megasonic], we make this kind of accreditation. They make a particle gradually out of the structure, right? So that's the one where we're positioned for that.

Second one, I can see that 3D, of the major applications down the road, a lot of multilayer interconnections and also 2.5 3D connectors, including TFTE for the 3D DRAM. So our covered 3D technology is very, very positioned for that. A recent we are making progress on the high-speed copper plating. And I think our key point is going to reach 3% for the right top-tier supplier. And that real attractive intention from the top tier customers.

And also another one as polishing. We talk about we can polish copper at a much less consumable costs. It's a saving of 80% compared to (inaudible). But another polishing tool that was for previous 3D side. And more than that, I should say, there's Furnace, right? We are very good there, getting to their traditional, I call it, (inaudible) and LPCVD, oxidation and all other corner process. But the real aiming for this product is ARD. We think Furnace ARD can play a very bigger role because of the efficiency, and for their future ARD projects.

So looking ahead, I mean our R&D is pretty target in the next few years, with the semiconductor manufacture trends. Further, we'll also try to develop potential new product right now, which we're going to announce very soon. Anyway, it's important for to see that ACM is -- technology differentiation is our core point. At this moment, we are heavily engaged with the U.S. and Taiwan customers. And we're in a portion of even giving demo for the key customer right now. I think the thing takes time, but we are confident this year we can penetrate 1 or more of the top tier customers, by our advanced cleaning side, also our advanced copper -- high-speed copper plating side. And that's our goal this year.

So Mark, anything you want to add on that?

Yes. The only thing I'd add, Charlie, is that this year, we wouldn't -- we didn't factor any revenue from these big customers -- from these new customers into this year's outlook.

Okay. Okay. Sure. And this question really is for Mark. I guess, a very -- 2 questions. First of all, the postponement of the IPO funding, that impact your funding for the new business or new trends, et cetera. Any impact to the operation? And how soon do you think that founding needs to be ready before that impact your business plan?

And second quick question to Mark is really about the OpEx ratio guidance or your OP margin guidance for the half of this year or for the coming 3 years.

Great. Great. Yes, I'll hit that. So on the China IPO timing, as David said, the progress is -- it's not under our control. So we're doing what we can to support. We're confident we'll complete it. Hopefully, it will be soon. But our business plan doesn't require an immediate IPO. And we talked about our CapEx plan, $10 million to $15 million, for this year. We would adjust it upwards after the IPO to spend more on Lingang.

And what we're doing, right, we feel like we've got plenty of room to add capacity in Chuansha. We're looking at other options there. But we feel quite well funded, and our plan doesn't require the immediate funds here.

In terms of the OpEx levels for 2021, David talked about the R&D. You got to think about our model, 40% to 45% on the gross margin. It kind of gets you to about a 14% op margin on a non-GAAP basis for the year.

Okay. So that means the OpEx ratio is around 30%?

That would be right. Yes. Yes.

Yes. Again, Charlie, we really balance growth versus visibility. At this moment, I rather spend a few points higher for R&D. And then we reduce a little bit profitable rate. That will be really good for our -- grabbing the market and also lead straight to our position. And that, I think, our philosophy behind it.

We have the next question from the line from Christian from Craig-Hallum Capital.

Great. Actually, the last few questions were the questions I had. So congratulations on a great outlook. I don't have any further questions to ask at this time.

Okay. Thank you.

Great.

Thank you. Seeing there are no more questions on the queue, I would like to hand the call back to our presenters for closing remarks.

Thank you, operator, and thank you all for participating on today's call and for your support. This concludes the call, and you may now disconnect it.

Okay. Thanks, everybody.

Thank you. Ladies and gentlemen, that does conclude the conference for today. Thank you for participating. You may all disconnect now. Thank you.