This call. Thanks, everyone, for joining today's Phison's 1Q 2021 Earnings Conference. Today, we're happy to have K.S., the Chairman of the company, he will begin with an introduction. After that, we will start the Q&A session. To raise a question, you can use the raise button in the Webex or enter your question in the chat room. Now, without further ado, K.S., please begin.
Yeah, thanks, [Ryan]. Good afternoon, ladies and gentlemen, this is K.S. I have prepared our Q1 result and some other Q&A collected from other investors. I will go through one by one. We just have only one hour, I need to go a little bit quickly and share more time for the Q&A. Okay. This is our statement. Anyway, please read. First page, we'd like, again, to re-emphasize that Phison, we made a lot of a big transformation from the consumer business to the technology-leading business. In these five years, we have a lot of activities, revenue, profit coming from non-consumer business. This is just to I have to keep repeating about this to brainwash your image about Phison. This chart will show you our Q1 revenue is still 30%, around 30% in consumer. Controller still maintain like 20%.
Something we have some big growth is embedded. Embedded, actually, when I read my report, internally, gradually by every quarter, we increase the number and also increase the share. The other one is about the gaming modules. Phison, we've been invest three years into the gaming business. We got some really good market share right now. In second half of this year, we keep working with the other two big guy. By CQ3, we are going to launch the other gaming applications. Please keep in mind. In industrial module, we have a very strong growth in the CQ1 this year, because last CQ2, a lot of developed countries, such as Western Europe, U.S. lockdown, a lot of deployment of equipment are postponing. Since the last CQ4 and the last CQ1, we got a lot of rush order.
Unfortunately, we are not able to fulfill 100% what they need, but we're still trying, and the demand on industrial is still very strong. Revenue, I think you know about our revenue is close to TWD 12.9 billion, and our gross profit by QoQ, we grow 28%. This is some business performance summarized. Controller keep maintain 20%. SSD controller, including SATA, PCIe, worldwide growth is 17%. One thing as our expectation, our PCIe controllers, worldwide we grew 96%. We have many design win with the semiconductor on the Gen4 controllers, which start to have a shipment and ramp up by CQ3 and CQ4. We still can expect a very strong growth in the PCIe controller Gen 3 and Gen4. [audio distortion] Hello? Can you please mute? Yes. Module. We have 74% are coming from modules.
One thing I need to highlight about at this March, they have a China gaming phone makers, Black Shark. They announced their very high-end smart gaming phone, which using Phison BGA SSD, and Phison, we are the sole supplier. We are, Phison, the first one to bring in their NVMe controller and modules into the Android phone. Because of that, recently, all the Android phone, except Samsung, now all working with Phison for the embedded BGA SSD into their gaming phone. We can expect by 2022, they have a very big growth on this business. They already have two smartphone makers that team members already stay in the Phison office to have our co-development on their new BGA SSD into their systems.
Industrial module, you can see gradually we are growing, looks like coming few quarters, the demand is still strong, but we are not able to fulfill what they need. In gaming SSD is a big growth, 39%, embedded ODM is 32%. We also able to see the embedded business is going to grow in the coming quarters. Overall, the revenue. Last year, CQ1, we are doing good because at that time the demand is strong, but unfortunately, the CQ2, a lot of lockdown, our revenue and margin dropping. Okay. Looks like this CQ1, we start to recover. Actually, since the last CQ4, we start to recover our revenue. Q1, we have a lot of backlog. We are not able to ship on time just because of supply chain is so tight.
CQ2, I think we have some breakthrough and the revenue already reflect on the April our announcement. Gross margin. Our peak was last CQ1. After lockdown, we suffer of market start to lower the demand, but looks like this Q1, we already recovered. Actually, the margin recover happened by March. By January and February shipment, a lot of order coming from last CQ4. Overall, the margin, because of CQ3 doing good, so we have a kind of a significant pickup at the Q1 overall result. I believe by our CQ2, the gross margin still able to defend in the high level. We made also some breakdown about our profit comparison between the last CQ1 and this CQ1. What is the difference? Basically, last CQ1 overall EPS is TWD 9.6, but we have a TWD 1.6 coming from the exercise of our investment.
Basically, this year, CQ1, the business in the operation is much better than the last CQ1. If we compare at the coming five, six quarters, seven quarters, our operating profit is a record high. Means we are getting the more good opportunity to make more profit to the company. This is some highlight here. Revenue is almost flat, QoQ and YoY. Inventory, you can see the inventory here. No, not yet. We have a good inventory management. The gross margin overall, CQ1, we have a big pickup compared to the last quarter, CQ4. Again, in our OpEx, we still have a high percentage into the R&D activity. We keep our commitment, then keep investing more into the R&D. Okay, so the gross margin, 29.5% is a high ranking in the record.
Not top ranking, but I believe potentially CQ2, we may hit the top ranking, the all-time high in the gross margin. This is the asset. I think you are more concerned about inventory. Since the last CQ1, market was good. Inventory is in a high level. It's TWD 12 billion. After lockdown, we start to manage the inventory. We lower to TWD 10 billion at the end of last year, but now we're back to TWD 13 billion. By end of April, it go to TWD 14 billion. The reason is we are not intentionally to increase our inventory, but we got so many orders, so we need to keep buying whatever the material, the power IC, the controllers, the wafer. By the way, the supply chain is very tight. Phison, we are not able to ship everything on time.
By April, revenue is lower my expectation just because we are not able to ship because of supply chain is so tight. In coming few months, the demand is still strong. I think we're able to maintain our high-level revenue. About Q1's business highlight, we award by the PC Gamer about the Best SSD for Gaming in 2021. In March, we also award by the Storage Newsletter the Best SSD Controller in 2021. We got the other award. This is just to prove Phison will invest huge R&D activities. We got a very good response. This also surprised Phison, actually. Last Sunday, Intel announced about their platform now is working on the Mars. On the picture, you can see Phison SSD is there.
This is just to prove it's a big evidence, Phison, we are doing good in engineering and doing good in the quality. This also helping Phison to make our good image to the worldwide Tier 1 player. Since that, we got so many good response, we also got some other new project approach from the key industrial supplier. This is surprise, but this is also is an honor to our R&D engineers. They did a good job. About Q1, our business highlight. Retail, actually, we have our worldwide first USB SSD controller, the single chip solution, we help our semicon customers to build 4 TB external SSD now already in the market. It's the highest density in the world. We also announced our SD 7.0 SD Express card now start to doing design- in with the notebook makers.
PC OEM, because of a supply shortage, Phison, we got good enough of controllers. We got many design win project with the notebook makers. We are working for three semicon with our PCIe Gen4 controllers start to ramp up. We can see more revenue coming from Gen4 controllers in the second half. Because of controller in eMMC shortage, we also got a few big Android phone makers and two semicon start to use Phison UFS controllers into the smartphone to replace the eMMC. We have no intention to design the high-end UFS controller because high-end UFS is only in the uMCP form. This dominate by three semicon who able to produce the DRAM. Two Korean, all 100% in-house. American DRAM makers, they use external controller house, but internal, they already have two controllers is ready, start to deliver samples.
Means for high-end model, eventually these three DRAM makers, they are only using their in-house controllers. Phison, what we did is we direct go to the smartphone makers. We propose the total solution to them with the middle low-end UFS discrete solution, and we got a very good response. Right now, we have to ramp up our UFS controllers mainly for middle low-end discrete UFS demands. We have two semiconductor working on this, and we have our China investor partner, and also we start to work with every smartphone makers. Also, we engage with the smartphone AP supplier. We build the ecosystem together with them. For sure, in coming years, in UFS middle low-end, Phison will play the position that today the eMMC. Phison able to provide the controllers and also the modules.
Our Gen4 new controllers, 12 nm start to have a sampling to the customers, and this will be our next growth in the Gen4 business. In gaming, we are launched the first BGA SSD into the gaming phone. Second half, we believe we got the other two, big approach design win is going to announce. Let's wait. We also, got the just announced that we have a PC Gamer that award the best controllers. Phison, by the way, in the coming summer season, we believe is a traditional, big demand in the gaming business. We are preparing a material for that. For server application, we also got a few design- in the China and U.S. market. Our new controller, 12 nm, is coming back by March.
We're able to deliver the high-end data center SSD by end of this year to the Tier 1 for testing. In industrial, because industrial now is so strong and we are on the Mars, recently we got again, so many inquiry. Automotive also is our Phison next approach. Industrial will maintain is a strong demand in Phison business. This Q2, our outlook. We start to ship our Gen4 value line SSD in the retail, not only Phison, but also with our partners, brand partner and the semiconductor partners. This will help Phison to get the more market share and revenue. PC OEM, Phison Gen4 working with the three semiconductor already in all the Tier 1 PC OEM under design- in. We also start to ship the controllers to them, and the ramp-up will happen by Q3 and Q4.
In gaming, we are working for the other two system makers, hopefully by CQ3, we are going to launch the first, the other one, system makers' ecosystems. Server is ongoing. Industrial, we are mainly focused to some of the security. We are shipping some very high-end security SSD to the U.S. dedicated agency. This is helping Phison to get the more market share and also making the much better profit margin. The outlook of semiconductor. Phison, our controller mainly in the 28 nm, 40 nm and 55 nm. Phison, we are different with other controller house. We also design our own SSD PMIC. When Phison selling a controllers or module, all with the Phison in-house PMIC. The 8-in is super tight. We don't see any release in coming two years.
Phison, we need to keep very strong relationship with our foundry, TSMC and UMC. We are also working very close with the packaging house to make sure when our controller flash is ready, we're able to directly go to the packaging and testing. Talking about the NAND flash, basically, since February, I believe the component shortage may cause the system makers start to release to downsize the forecast. Looks like this happens. Smartphones start to downsize, the notebook start to downsize their forecast, right? By February, personally, I think the flash may going to solve by CQ3. Just months ago, the Chia coin, the cryptocurrency use a storage start to getting crazy. Now, I believe CQ3, the flash will get tight again. From the informal motivation from the semiconductor, they are going to raise their NAND flash price around 10% by CQ3.
This is unofficially. We are waiting and monitoring. Phison, we keep investing very strong activity in R&D. Enterprise and security and recently Chia. We are not like to focus to Chia, Phison at least we need to figure out what kind of SSD algorithm is the best for Chia. Last Friday, we found some good algorithm. We are going to have some white paper to our customers. With this kind of algorithm, we're able to create a much better profit on the SSD to Phison. Phison in our Q1, we still maintain very strong investment with R&D. This is our commitment to make sure every quarter, Phison able to deliver the new products to the market and to the customers. Our headcount, you can see by end 2020, we have 1,500. Now we are already going to 16.
By end of this quarter, we are going to 17 or 18. We still need a lot of engineers because so many projects are queuing there. Phison's strategy, we have defined the so-called five plus five. The left five is a Phison business model. We have IP, controllers, modules, design service, and full turnkey. This is our business model which able to create much better gross profit and net profit to the Phison. The business section, we have these five: gaming, enterprise, automotive, industrial, embedded. Every one of these business can become independent company. This is big enough in the revenue. Phison, we wouldn't spin off anything. We just use our business model to create more revenue and profit to the company. This is our goal and our strategy. We collect some Q&A which I'd like to make a brief report here.
The Chia cryptocurrency actually making a lot of, maybe you can say, noise or demand in the market, and we get a strong demand, but we are not going to fulfill what they need. We need to use our resources to fulfill first to our designing projects, such as a smartphone, TV, notebook, auto. On the other hand, Phison needs still to study what kind of algorithm is good to the Chia cryptocurrency. Now we found something good and we are going to announce to work with somebody. At least this is helping Phison to sell more SSD and create a much better gross margin. Looks like controller supply is there, but our demand is getting too strong. Yes, if you ask me, our supply-demand is getting worse because getting so many projects.
I visit to UMC and TSMC recently, we try to ask some more upside support by CQ3 and CQ4. If there were some downsides in the forecast in the phone or notebook or other application, this definitely will help Phison to get more resources in the foundry. By the way, if Phison, we are only in the control business. If you can know how many wafer are able to get, then you can know my revenue is there. Phison, we get the certain of a controller wafer, we able to allocate the wafer, go to controller business or go to the modules. If go to the module, for example, 2 TB SSD was like $200 FOB from our company. Because of Chia, we making some new algorithm, we able to sell the drive up to $300 with the same materials, different technology, different applications.
If I'm only selling controller, I can raise controller from TWD 5 to TWD 10, but just only TWD 5 extra income. We build in the modules, we gain extra TWD 100 revenue with TWD 100 gross profit. Again, this is our business model, and we able to create the net profit to the company. Controller shortage actually is good to our module business. This is asking revenue is the same, why EPS is same as last year. I need to highlight, look to the last CQ1 versus CQ1, the revenue is almost flat. Actually, the last CQ1, the flash was in shortage. Unfortunately, because of CQ2 drop down by COVID-19, so revenue dropping gradually back. We look to the margin, last CQ1 was at 30%, and this CQ1 almost close to the 30%, means a flat.
The difference is CQ2 versus last CQ2, personally, I believe there were big growth in the revenue and big growth in the gross margin and also in the gross profit. Okay? I believe by CQ2 and CQ3, we still stand in a very strong demand. With our business model, we able to adjust our portfolio to make more profit to the company, then invest to the R&D and also return to the shareholders. Wafer is tight. We keep very good relationship. We create a value to both TSMC and UMC. They recognize Phison. We have a very unique position in the market. Then they are more willing to support to Phison. About testing packaging, we already react since the last November. We ask the packaging house to increase the price.
We talk to the substrate, we able to make the prepayment, we able to give them some price increase. In packaging and testing, I think now we are doing good. Only problem is wafer is still not enough. Even though we got the upside, we are still not enough. Phison, our business, we already diversified a lot of business to the different portfolios, such as auto, gaming, embedded, industrial. Basically, I don't worry if the cycle go to the downtrend. I believe if IT market go to downtrend, everybody suffer. Of course, Phison suffer, but I think we just only minor suffer because of different portfolio, and we able to manage. I don't much worry, and I wish the downtrend come as soon as possible. Because during downtrend, Phison, with our business model, our strong finance, we are able to react to the market.
For example, last year, even downtrend happened by Q2, Q3, we still able to grow the revenue and making much better profit. I don't mind the downtrend coming because we able to manage with our business model. Okay. How we react, I think this is our portfolio. We have so many different portfolio, so many different business model, I don't mind and I don't worry about it. Supply-demand, basically, flash Q3 originally I thought will become soft, because of cryptocurrency Chia, I believe by Q3 the NAND will be too strong. Already three semicon announced they are going to raise the price on the NAND. That we wait. Phison have our strategy in China. Last year, we put our subsidiary merge with our JV company.
The reason is we need to use a China entity to get a China business, and this is doing good, and I can see they are expanding and getting more market share in China. Phison is one of our major shareholders. We are able to gain, not just only from business, but also from capital gain in the future. Colorado is mainly for the enterprise SSD controllers. We keep investing, and the return will happen only after two years because this is a long-term investment. What is the Phison vision? Again, this is our five plus five. Left is the business model, right is the target market. Each market actually can be independent company and can be big growth. Phison will cover everything inside the Phison, and we are not going to make the spin-off.
Within five years, I believe every segment will be able to grow, to help Phison get more revenue, market share, and also profit. This is about the ESG ranking. To be honest, in the past, Phison, we are not focused on ESG. We don't have a dedicated department. Since last year, we start to invest, organize a good team. Now in Taiwan, OTC ranking, Phison now already in the top six to 20%. Also highlighted by OTC, we have a big improvement. MSCI ranking was CCC because we did nothing. Just last year, we start to getting aggressive, prepare the document. Now we already in BB. I can see very soon we will be upgraded again. This is others, all the accumulated question, a ny, we can go to the live.
Thanks, K.S. For investors online, please use the Raise Hand button or enter your question in the chat room should you have any questions. If not, maybe I will start a few. First of all, thanks, K.S. Maybe could you provide some updates on the flash and the controller price trend for coming quarters? After the strong margin performance in first quarter, what do you expect for the coming quarters?
Actually, Phison shipment January, February, which we ship the order from last CQ4, with the relative low profit margin. The effect on the margin will start from March. We also have a lot of backlog we are not able to ship on time by March and April. I can see in CQ2, we are going to be strong in the revenue and also the gross profit. Talking about CQ3, I believe, even though they have some adjustment in the system house because of material shortage, Phison, again, we are so many portfolio in the business and also the crypto Chia is helping us to get some other new business. I believe CQ2 will be strong and no doubt in revenue, we are able to hit the record high, the all-time high in CQ2, and profit, gross profit, potentially, also may able to hit the all-time high.
Q3, I still believe the demand is still strong. Coming few months, I think we need to work very hard to get more supply from TSMC and UMC. Yeah.
Yes. Thank you. We have a few questions from the chat room. First thing is Kingston stake is no longer shows in the 2020 annual report, and Simon is asking why is that? The second question is, the impact of Intel's new 10 nanometer CPU to support PCIe Gen4. Also what's the implication from Silicon Motion's published guidance?
Phison already sold back the KSI share. Hello? Hello, Simon. You have an echo again. Let me answer one by one, okay? Last Q3, Phison already sold back the KSI share to the Kingston, I think you got this kind of data, right? By law, we are not able to show any report. Phison has nothing to do with Kingston. They are just our customers, partnership. We work very close together, co-development, we are not in JV anymore. Second question about the Intel CPU. I not much worry because Phison, our allocation in the wafer for Q4 sold out. This year sold out. We are trying to ask more supply from TSMC. If I able to get more supply, then I may have a concern on Intel, at this moment, sold out.
I think I have no any concern on that. About the Silicon Motion published guidance, actually, you may need to ask their customers, the share of their customers. If they have a customer from, let's say, from 10% up to 50%, then you have to ask why, and what's going to happen. Simon, please. Okay. You may need to ask. My understanding, there is a UFS controller ramping up because of the ecosystem, but Micron also has internal two controller ready. If a customer occupy like 40%, 30%, 50% of the revenue share recently, then you have to ask, how about next year? Okay? This is my comment, and I have no any interesting to answer where is the published, but I only looking to Phison, our business, how we able to make more money. Okay?
Thanks, K.S. Simon, do you have follow-up?
Yes. Can you hear me?
Yes.
Oh, yeah. You know what? All the background of the Webex is Chinese language, so not easy to find how to raise the hands or raise the questions.
Yeah
-maybe next time you guys can convert Chinese background of the Webex into maybe English, if you don't mind. Anyway, yeah, thank you very much. Very quick is, when we look at the NAND flash industry, the industry consensus indicates about 35% big growth there will maybe price recovery. How do you appear the NAND industry cycle the implication for Phison? The simple question is, how do you assess over Phison's timeline growth for second quarter over 2021? The, yes, you said gross margin not really improving year-on-year, any color for the gross margin for second quarter and 2021? Thank you, sir.
Okay. Let's go this way. Phison, we are not a NAND manufacturer. We're just a NAND users. We don't mind the NAND manufacturer to keep investing more, making more NAND, making NAND more cheapest. Again, due to the Phison's, our business segment, we are going to use more NAND flash in the coming future. We wish the supply getting more, the price getting better. Hello. Simon, you have to mute. Okay. I have no any comment on how the supplier to increase their output, but we are happy to see they have more output. Look to here, Phison, we have over 74% business are coming from module. If the NAND flash supply is more than enough, price is much better or lower, we are more than happy. We able to create more systems in storage to support to our customers.
Phison, we have two main businesses, controllers and modules, right? Controller is good, it's important, I'm still always a question, how controller company able to make the more profit from controllers since you have a big investment there? For the mainstream controllers, semicon, most likely they want to use their in-house. For example, the uMCP are dominated by three DRAM makers, two Korean, one American. Two Korean is 100% in-house controllers UFS. American, they also have two controllers UFS ready. If we rely on that, I think I'll be killed in the next few years. Good thing is, Phison, we have so many self-control. I define a so-called self-control module business. We make the customization, we make a lot of new different applications, create the value that able to help Phison get the more profit.
I want to share with you, recently, in the last year, we start to co-develop with the U.S. agency. We create a cybersecurity SSD, same material, 2 TB in market, $200, but with support to the cybersecurity, that can be $350. Same material, different price model. This is the things, Phison, we are going to build our own ecosystem, our self-control business that we're happy to see the NAND-Supply bit increase every year. Okay?
Thanks, K.S. Also another question from Jeff. How much and how many times was your controller price raised this year?
Okay. Phison, we design every controllers. You may not know, Phison also developed the power IC PMIC for our controllers. Let's talk about PMIC 8-in. We've been raised three times from the supplier in the wafer, 8-in wafer, three times. We also reflect our PMIC price three times to the customers. The price already over 100%, because our PMIC is so tight. Okay. Talking about their controllers, like USB memory card in the past, the price is too low. Those price also already reflect 60%-80% in the market, and we raised two, three times. The most tight in controller is the eMMC. We not able to raise so many times, but we also increased the price around 40%-60% to reflect the tight in supply. SSD, we also raised two times because, again, demand too strong.
We're not able to get good enough supply. Overall, we don't like to raise the price in controller, but we have to because the wafer price, packaging price, testing price was increased from our suppliers. Okay?
Okay. Thanks, K.S. Maybe one question from me. For the coming quarters, as you mentioned, second quarter is likely to be a record quarter. Probably by end applications, what do you expect to deliver a much stronger growth in coming quarters, and which segments could be below your average?
Actually, this and next quarter, industrial was strong. Embedded, super strong. Gaming, not yet. Gaming is still flat, but I believe next quarter, Q3 gaming will be very strong. Controllers is always strong, but we are not able to fulfill the demand because of tight. We still need to keep some controller inventory for our in-house applications. The next is about the crypto Chia. This is strong, but we still need to think how to react on this extra demand. Yeah.
Okay. Thank you.
I saw Simon has a question, but I'm not able to read here.
Oh, okay. He has a follow-up on the previous question on ASP. He's confused with the 1Q ASP for controller and the non-controller products. Could you probably recap 1Q price trend in terms of QoQ?
I think with 1Q ASP high for controller based on this.
Q1 controller price in the USB memory card, almost 80%. PMIC almost raised 100%. SSD controller SATA 10, 20, 30, yeah, they also happened. The problem is, a SATA controller was at $2. We raised to the $2.50, means the raise 25%. It's good. We got extra 50% margin. On the other hand, packaging cost is increasing. We use the controllers to build a module, for example, 240 GB SSD was $22, $23. Because controller shortage, we're able to put the drive to the $28. Phison, I can either sell a controller, get extra $0.50 of margin, or I decide to build the modules to get extra $6 of margin. I'm smart, right?
I think Q1, we are trying to take the more controllers into the module business to help customer to get the more modules. On the other hand, for designing controllers such as our partner Kioxia, Kingston, Micron, WD, YMTC, we keep supplying the control to them. Even though the price increased just little, we have to loyal to our commitment. Phison able to make the more margin from the module business. Okay? Okay. The margin improved so little, to be honest. Phison January shipment, which follow Q4 price. February shipment follow their December price. March shipment follow the January price. From order to shipment, we need to have 30 to 60 days. March, we have a much better margin. On the same time, by April, May, June, I think we are in the high level gross margin.
PCIe is, we have a 12 nm and a 28 nm, mainly in the TSMC. Okay?
One follow-up from me as well. As you improve your gross margin this much, you also have progressive R&D investment plan in coming quarters. In your slide, we saw that your R&D headcount serves quite a bit in first quarter. Maybe could you share your target for the year? Also if there's any change to your R&D headcount target by 2025, as you mentioned in prior quarters.
R&D headcount to me, no ceiling. Until 2023, no ceiling, because we still have so many project shooting. We've just been requested by one of the Korean semicon to us develop controllers. We're also asked by the U.S. company to do the other controllers. Japanese also asked, but unfortunately we don't have a resource. So many project we have to keep on hand and try to find a resource. My goal is, I hope if I'm able to get enough engineers by end of this year, we need to hit 2,000 in engineers. By 2023, we wish to have 2,500 to 3,000 of engineers. This is a heavy expense, but Phison business model, we're able to create more than enough profit to subsidize this R&D activity.
Okay, thanks. One follow-up. Given that R&D headcount plan, how does that imply to your R&D budget, in terms of either absolute terms or as a percentage of revenue?
Now we still keep 12% to 14% of total R&D activity. We are going to keep this kind of a number. When our profit getting better, we are going to inject more. I think 12 ± of total expense is our target.
Okay, thanks.
Simon have some in the writing, let me check. Auto NAND spec and also auto NAND controller, please. Today's auto, the NAND in the automotive is relatively small because it's only 16 GB, 32 GB, 64 GB. Imagine your USB drive already 128 GB. A car using less storage than your removable storage. This business is good, but too small in the ASP. Phison, we have a high-level enterprise-level SSD controller modules for the auto drive application. This only will happen in 2022 or 2023. We are investing, developing right now, but we are no rush. When times happen, we are there. We are standby. Yeah. Great. Yes.
Okay, K.S., one more question from Simon, also asking YMTC China NAND flash sourcing.
YMTC still is a Phison main supplier. We are working very hard for their 128-layer and start to ramping up from coming June. They are our very important supplier. Phison control to YMTC China meaningful? Yes. Why not? China keep looking for their own products, their own design, and looks like YMTC flash is proven in technology, also proven in quality. It's meaningful, and Phison we have a JV in China that is going to support more into YMTC business. By the way, I think no doubt China is the biggest NAND consumption in the world. We have to do something to make sure we are getting more share in China.
Thanks, K.S.
CQ2, yes, sales definitely in the all-time high. Gross margin, I believe, also in all-time high. If COVID-19 no more impact to the market, okay? If. COVID is not we can control, now look to the May, June, still very strong. Phison, we don't provide the guidance because of our history and Phison, our business model. I think April, we are in the TWD 5 billion in revenue. Overall, we believe we able to keep flat or growth little bit. This depends on the supply chain. We still tight in the some flash memory. We also tight in the controller. Simon, can you hear me? You may can try to unmute to see if I still have a echo.
Okay. How about now, sir?
Now it's good.
Yeah. We really appreciate your great presentation and also all your answers to the questions. Maybe almost the last question is, your foundry sourcing, you've been saying UMC is a great company which will make continuously your controller, and meanwhile, your competitors will use continuously the TSMC. The question is, TSMC is a very big foundry company versus UMC, relatively much, much smaller than TSMC. That really means some piece of advantage for Phison. They're acquiring some new capacity. The question is, to promote the PCIe fourth generation controller or more advanced, do you need, maybe what? Do you need a five-node, 14-node, 16-node, or still okay with the 28 nm-node, and also the UMC's capacity is still big enough for your double-digit strong growth? Thank you, sir.
Simon, I think you didn't make a close follow-up about Phison's portfolio. Phison's our 8-in, 55 nm, 12-in, 40 nm, 12-in, are in the UMC. Our NVMe, 100% in the TSMC, 28 nm and 12 nm. Since last year, this year, we have more than 20,000 pieces of wafer support from TSMC. If talking about 28 nm , TSMC is our biggest supplier. Hopefully by next year, 2022, we're able to grow to 40,000 to 50,000 pieces per year. We cover both UMC and TSMC. Okay?
Thanks, K.S. Maybe one follow-up from me. Regarding your foundry capacity, when comparing 2021 versus last year, how much do you increase your foundry capacity support?
We give our forecast to the foundry, both UMC, TSMC. Unfortunately, they're still not able to fulfill what we need. Again, now it's just May in 2021, still too early to talk about 2022 whole year supply. I deliver very important message to both TSMC, UMC executive, that Phison, we are doing a lot of activity in the design win. They're able to rate all our achievement every month. They can make the relative adjustment anytime. I'm still quite optimistic. We are going to double our wafer supply next year. Although at this moment, we not get 100% commitment. With time flying, we get the more design win, we get the more award, and also from the approval from our customers, Tier 1, I believe we are able to get more wafer supply.
This all depends on Phison, our R&D activity, and also our position in the market. If they agree we have a position, I don't much worry about our supply next year.
Yeah, thanks. Probably still too early to tell the 2022 number, but how about this year versus last year, for the parts that are being committed already by foundry partners?
Basically, right now, their foundry, what they do is next year flat as this year. This is the principle, depends on each different company, right? I think we able to get some more because of different, our position. We also making a lot of ecosystem partners with the tier one chip makers, and they also need Phison. They will talk to foundry to support us. I don't worry too much about it.
Okay, thanks, K.S. Also one question on the overall smartphone market, from Randy. Randy's asking, the Android market seems to have more news on softness, some supply related, but also demand seems mixed. Are you seeing any softness in this mainstream controller for that factor?
Actually, since last year when Huawei started to quit the smartphone market, those few big China phone makers already gave a forecast in this year, which is 2x-3x of Huawei's total demand. Last CQ3, CQ4, we already know they are going to make adjustments. This happened, right? I not much worry about that because our controller is still not able to fulfill. Even though they adjust, we still not able to fulfill what they need. Okay?
Thanks, K.S. Randy, do you have a follow-up? Maybe one from me. Still in the smartphone related market. You already have a project win and shipment to Xiaomi on your BGA SSD. Could you update us with some of your other projects regarding the BGA SSD in coming quarters?
All the Android phone maker, except the Korean phone makers, already approach Phison. We have two customers now start working in Phison office to co-develop the driver of the Android phone BGA drive. We believe next year, Phison will cover most of the high-end Android phone, gaming phone market, and looks like we are the key supplier there.
Okay. If we compare that BGA SSD solution, what's the, say, pricing or margin difference versus other product lines?
This is a proprietary customization product where the margin is better than the module business. Better than any other module business, because this is a customized product.
I see. Is there any estimate or expectation on the smartphone SSD contribution revenue probably this year or next year in terms of percentage or Yeah?
Okay. We got the forecast from the five phone makers, which now doing the design- in for next year. Overall, again, this is so-called forecast. Okay. Is eight to nine million pieces. Eight to nine million pieces in a phone is small, but eight to nine million pieces SSD to us is big enough. I don't know if that's true or not, but at least Phison as a sole or major supplier, I think we're able to get the most of the market.
Okay, thanks. We have a few questions from chat room, also from Simon. First is on the dividend policy and the CapEx trend, this year, next year.
Dividend policy as a pass, we keep 55% of net, our earnings per share, 55%-60%. Okay. Again, Phison, we are going to keep hiring more engineers. We are going to build more building to that engineer working. We are still growing. Right now, our revenue is close to TWD 2 billion. My goal is to go to TWD 3.5 billion within two to three years. We are going to increase the inventory, we are going to increase the business, we are going to increase headcount. Company need to keep the cash. For many years, Phison didn't raise capital from the market. If need, Phison may go to get the money from market. If through the earnings, we prefer to keep 40%-50% on hand, at least 40% on hand to grow our business. The effective tax rate, the CapEx.
Phison, we don't have a factory, so we don't define the CapEx. There's the OpEx. The only thing is a new building. Every two, three years, we are going to build a new building. Is not that much number, still affordable. We don't have a CapEx. We only got OpEx. Effective tax rate. Taiwan tax, I think now is at 20%. We still have some incentive. I think overall, we are in the 15%. One, five. Okay?
Okay. Also, one more question is asking, is there any negative impact from OEM's production cutback? Meaning probably component shortages, that leads to the cuts back by OEMs.
Again, since February, I already talked to my people. They have adjustment in the inventory from system makers. Now looks like happen, right? We already prepare for that. Even though they made adjustment, we're still not enough to supply. I have no worry on that. Yeah. Okay.
Okay. For investors online, should you have any further questions, please enter in the chat room or just unmute.
Okay.
Okay. If not, probably, I think, K.S., we can wrap up the call.
Yep. Thank you. Overall, I think our CQ1 result, to be honest, the profit is over my personal expectation because January, February, we are shipping most goods followed by last CQ4 orders. This is over expectation, means we got some out perform performance, the out perform the profit, so we decide to increase the R&D expense. CQ2, I believe we are in a very good shape in the revenue and profit overall, and also the some market share. Now what we have to do is to working hard to ask both TSMC, UMC to give us more supply wafer in Q3, Q4, helping us to fulfill the demand from our customers. Every quarter, Phison will give a new surprise to the investor that, CQ1, we have BGA into the phone. Now we have on the Mars.
CQ3, definitely we are going to have some good news on the gaming business. Please keep waiting. We have a commitment here. We are keep investing more in technology and at all, the return to shareholder is a much better profit. Okay. Thank you for your time today.
Thanks, K.S. Also thank you for your participation, and today's call end here. Have a good day.
Thank you. Bye bye.