Good afternoon or good morning. This is Charlie Chan, I am Morgan Stanley ’s semiconductor analyst. Today, we are very pleased to host this second quarter results meeting for Phison. We're very happy to have our company's chairman, Mr. K.S . Pua, with us to share their business outlook with us. For today's event, I will start with the company's briefing and then the Q&A session. In any time of this webcast, please type your question in the Ask Question text box. I will later on, will moderate the Q&A session for you. Without any further ado, let's welcome Mr. Pua to share the second quarter results and third quarter outlook with us. Thank you.
Thanks, Charlie. Thanks, Morgan Stanley. Good morning, good afternoon, ladies and gentlemen. This is K.S.. Let me start today’s my presentation, then follow with the Q&A. Let me share my file. Okay. Yeah, this is our statement. Take a few seconds to read, then we directly go to today's presentation. First section, I would like to, again, to share the Phison business transformation. For so many years, they have a lot of investors keep in their brand. That Phison is in consumer business. The answer was yes, many years ago, but in these few years, Phison made a big transformation from their consumer market into their embedded design in customization and storage business. 2020, Phison have a big growth in our revenue. The reason is, we increased lots of portfolio such as gaming, embedded system.
We also grabbed that this first half 2021, we hit $1 billion in the half-year business. Again, a lot of business are coming from non-consumer business, so please have a new image about Phison. This is Phison's, our business portfolio. Consumer, last quarter was 30%, now it's going to 26%. Controller, we can maintain that 20%+. The embedded ODM is 14%. By revenue-wise, this business two years ago was nothing, now we can see rapidly increased amount. Gaming module, we also increased this portfolio since last year, and looks like this is a very strong demand, and we play a very unique position in this business. Industrial module, since actually last year, Q2, when COVID spreading to the Western country, the industrial demand suddenly frozen.
Starting from Q4, the demand was good and still in the big undersupply at this moment. Around we have a 70% of business are coming from non-consumer business, and more are related to the high investment in R&D and the highly customizations to the customers. This morning we just approved by board members about our Q2's financial result. Revenue-wise, we hit close to TWD 16 billion. QoQ was 23%, and we made a record high in revenue. The gross profit, TWD 5.1 billion, which QoQ growth of 36%. Also, we hit the record high in the gross margin. The reason of revenue, because of a lot of controller and our modules, our products under supply. Also, as I mentioned, we have a big portfolio in the high-margin business, which is all the quest for design-in process.
By margin-wise, I think we are good, because of, again, from a different portfolio. We also got a lot of design win project with the key customers and the key players. This is some highlight here. Overall, by Q2, 20% revenue are coming from controllers. The SSD controller worldwide growth is 50%, but PCIe Gen4 plus Gen3, the growth is almost 200%. As two years ago, we start to invest a lot of resource into the PCIe Gen4, now we start to see the return. By module-wise, we have a 74% of business are coming from modules. Again, these modules is not consumer, okay? It's embedded design in industrial and also gaming and automotive, which able to help Phison to increase our influence power in this industrial. We also have a new product with so-called BGA SSD.
Mainly, we ship to the automotive, gaming, and some high-end Chromebook, high-end notebook, and also from some Japan gaming systems. Which we also can see the very strong demand and the growth is 28% worldwide. By industrial, because last Q2 the market was almost frozen. Unfortunately, the supply is under constraint, including controller and flash are in the tight supply. We keep seeing very strong demand coming from industrial application. Just one thing to share with you. We made the design win into the one of very key robo makers in the Japan. Our system means controller with the flash modules. We have to commit to supply for 30 years. 30 years. Anyway, this is a kind of a niche in this market, but anyway, for 30 years, of course, we have to commit, but we try to find solution to commit. Okay.
Again, this is a good business. By gaming, we got a big growth. You can see in this two years, gaming start to adopt the SSD to replace hardware, and this will be our main business. By the ODM, the growth is 54%, and quarter-by-quarter, we can see a very strong growth. Revenue, hit the record high close to TWD 16 billion, and we have a good YoY. The gross profit, gross margin is 32.5%, and also we are in the all-time high. The reason is because of supply constraint, demand was strong, and also we have a lot of new design with a high margin. Earnings per share by Q2 is close to TWD 11.5. The first half this year is around TWD 20 per share EPS. Again, it's also all-time high by quarter.
Almost all the EPS earning are coming from our business. This is some highlight in the statement. Revenue-wise, because of U.S. tend to weak TWD, new TWD, the TWD dollar is getting strong. By count to the TWD, the growth is not that much, but by dollars, actually, the YoY is 56%. Okay. QoQ is at 24%. Gross profit YoY is 93%. I think, to me, is under my expectation, but also I think we hit a good result to the employee and also return to the shareholders. The OP, YoY is 134%. The net income is a YoY 91%, QoQ 34%. Overall, first half this year, EPS is TWD. We hit the other all-time high, and I'm happy and pleased, appreciate to my staff, my people, they are working so hard.
Especially by the May and June, Taiwan was attacked by the COVID. We have over seven weeks, only 25% people come back to office. Like 75% are working from home. A lot of schedule communication create a lot of problem. Anyway, my people working hard, we overcome the limitation. We hit the good business, I think this is our commitment to the shareholders. Okay. Balance sheet. I think you are more concerned about inventory. When you look to the inventory, last year was TWD 13 billion. This year, by end of June is TWD 15 billion. The reason we increased the inventory because of a lot of material, the supply. Originally, like four cases, lead times are four weeks, now 10-40 weeks. Foundry is very tight. A lot of PCB capacity, active component from four to eight weeks going to 30-40 weeks.
We have to increase our inventory level. On the other hand, we missed so many design win into the system, especially for the industrial project, which they ask, we have to supply for three to five years. If the component is going to EOL, we need to have a last buy to keep five years consumption. This also increased our inventory. Increased inventory, I don't feel this is a heavy pressure because those inventory most likely already have a business. We're not able to ship by tomorrow. We have to ship this month by month and then continues for few years. Inventory increase is helping Phison to gain the more revenue and also increase our gross profit. Okay. The net value per share is TWD 174 because we just have a dividend, TWD 23 will happen by next Tuesday. Okay.
The ROE is still able to maintain the 20%+. Today we also announced our July. Our consolidated revenue is TWD 5.4 billion, close to TWD 5.5 billion. We still under impact by the COVID in May and June. Our supply chain is still under recovery. We still have many products, ICs need for testing, still under queuing. We hire employee, we hire the tester. Phison will pay to get the people and send the people to the factory to support our supplier to test our chips. I believe by end of August, everything will be back to normal. Of course, we got a lot of complaint from our customers from June to last month. Anyway, after September, everything will be back to normal.
Revenue close to TWD 5.5 billion and YoY is 42%, we can see the net demand market still maintains strong. We still need to work hard, try to get more business by this year. Okay. Phison public for 16 years. This morning, board members approved we are going to issue the first domestic un secure d convertible bond. You may surprise why we need to have a CB. This total amount is close to TWD 3.5 billion. The reason we have to apply to the competent authority to get the approval. The reason for this issue, the CB, you already aware the foundry is very tight 2022. Still, we're not able to get full supply 2023, no idea yet, but we need to secure more supply in coming 10 years. We have LTA with our foundry partner.
We need to pay the deposit to them for many years. We need to collect some capital to make sure we able to ensure our long-term supply in the wafer, mainly in the 55, 40 and 20 nm. Phison, we need to expand our business scale. You are aware that the NAND market still growing. Phison, we cover IP design, service business, SSD business, and also the module business. Which module business may need more capital. Module business helping Phison to get a more market share, more revenue actually, and more income. We need to get more capital to help us gain the business. This first half, Phison, we increase up to close to 350 headcount in R&D. The whole years looks like we're able to hit the 800+ new headcount in R&D.
We need to buy more equipment, tester equipment, developing equipment for new people, new engineers. That's the reason why we need to collect the capital through the unsecured CB. Our goal is to get the new capital, we need to build the highest walls in the technology, getting the new people, enable more projects, getting the more design win. This is to ensure Phison get the more return in business in coming years. Based on our experience, if we invest the new team members, after two years, we can start to get the return from the business. Also we need to get the more resources because we have the other 5+ 5 Growth Strategy, which we need to cover this five segment business and to get the more return from this industry. To this, board member also approved.
By end July, Phison shareholder meeting approved the dividend. In the past, the dividend only happened once per year, but after shareholder meeting approved that the dividend will go to 2x per year. Our policy will be, every Q1, we are going to pay last year first half's earning. By Q3, paying the second half earning. Approved, TWD 10 per share dividend will happen by next year February, to the shareholders. When we decide policy, then we execute. From now on, Phison is going to pay 2x per year. Okay. Some business highlight to share with you. This is by Q2. Our Gen4, the fastest, PCIe Gen4 modules, we start to ship to the market, but because of a controller constraint, we are not able to get the more shipping. But gradually, we are solving this controller problem.
So I can see these products will go into the market as soon as we can. W e have two NAND makers announce the PCIe Gen4 SSD with our Phison, the new controller solution. Now start to make the design in testing qualification by the top five of PC makers. We have external drive up to 4 TB to our semiconductor partners now in the market, and also our finished products, ODM to our storage customers. OEM-wise, our Gen3 controllers, we got design win and shipping to the Chromebook, high-end Chromebook, tier one customers. Again, the Gen4 start to get all tier one OEM design in qualifications. By gaming, we already also got the design win into the cloud gaming platform. This is the so-called data center. We start to ship the products and ramp up by next year Q1.
We also enable the new highly customization gaming notebook with our high-end controllers. This is the highest performance. This product will go to the market by end of this year. Also we're putting our SSD solution into the portable gaming phone. Okay? By data center solution, we keep investing a lots of resources. Gradually, we can see getting some design win project. If talking to the revenue-wise, I think this will happen only after 2023 because this is a long way to go and long investment, but we never give up and we keep investing. Industrial, as I mentioned, is a too strong demand, and we are not able to fulfill the demand. Specialized, I want to talk about the security products. A lot of countries' governments start to realize storage security is important.
We have a partner with our U.S. customer, Cigent, to make very unique SSD for cybersecurity, which start to sampling under qualification and start to ramp up by Q4 this year. Also we got the design win into the robo company in Japan, the major robo company, which they're asking for 20, 30 years supply. Q3 business outlook. Retail was slow because of several countries still are under trouble by COVID. The Western country, developed country now people are enjoying the summer vacation. We believe the retail business will be back by September, October. This now we are preparing to get those business. For OEM, we keep design in our Gen4 controllers with our semicon partners, with our modules partners, and also with our modules to the all tier of the PC makers.
In gaming, recently, the PS5 is announced they are going to let the user to extend their SSD solution. In Phison, we are on the list, we are supported. Server-wise, again, keep a long way to go, keep investment. Our high-end silicon Gen4x4 controllers is silicon-ready. We need to submit the samples to our customer by Q4. My partner will deliver this to the tier one data center company by next Q1. I believe revenue-wise will start to happen on the business by 2023. They will take one year for qualifications. Once Phison is ready after 2023, data center server business will become our main business. Again, industrial, a lot of design in, design win, customization keep asking. This will need a lot of R&D headcount. Q3, they may have some special underscore light.
I think recently the news keep talking about the Sony PS5, and they have a slot. They are going to upgrade their software to enable this slot. Phison, we have a few partners already made announcement. The products are qualified and confirmed, are going to apply into this business. We can see some new business coming by very soon. When Sony starts to release the upgrade, then we're able to get the business. This Q2, we put our SSD into gaming phone, and that gaming phone become the fastest performance in the world. We also get some design win business into the handheld gaming equipment. Sooner or later you can see the products in the market. Again, by the security now is very strong requesting, and every country keep asking the different design.
Phison, we accumulate over 10 years experience on the security. We believe this is a highly requirement in developing and also high mix, small volume, but definitely is a super high ASP and margin, and this will become our cash cow by 2022. Overall, what I can define as a suffer about foundry supply, we still far than enough. We wish our partner able to gradually increase the wafer supply to us, that we spend those huge amount of money, people in the new business. We need to have more supply to fulfill our customers. With the LTA here with the foundry, I think this is to secure coming years of supply. Again, still not enough. The NAND controller is in the constraint. First of all, we need to fulfill our customer's demand.
We need to make sure our customer, once they get design win, they're able to get controllers. Also, Phison will try to get our own controller to build to the modules to increase our ASP and also our gross profit. We need to be very smart on this business management. Anyway, Phison keep working hard, increase our R&D people, increase the equipment, keep hiring the more people. The goal is that we need to prepare for coming TWD 100 billion revenue. This is our goal, and I wish this happens soon. R&D headcount, first half we got extra 300, but whole years looks like we're able to get over 100 headcount. Space is not enough. We need to pull in our facility and also to keep buying more of the equipment for our new headcount.
The goal is to have 2,500 by 2025, but I believe by 2022 second half, we already hit the 2,500 headcount of R&D. The reason we keep asking to get more people, because semicon, our partners, keep asking Phison develop more solution for them, which including the client, data center, automotive, industrial, security, imaging. I can't mention because all the systems in the NAND flash and those NAND flash solution most likely went to the semicon but also come to Phison. That's why we need the more people. Okay. We collect some Q&A. We finish with Q&A, then we go to online. Investors keep asking what's the impact of a weak USD to Phison. Anyway, we have finance prepare our explanation here. Overall, Phison, our material mainly in U.S. dollars and our business also happen by U.S. dollars.
We are natural hedge. By Q2, the foreign exchange loss on the book is around $57 million, which is 0.36% of revenue. I think we are making the good management on this, and the damage is well under control. We wish this we are able to balance by coming few quarters. Weak dollars impact to our revenue, yes. Profit, yes. Not that big damage compared to some other business. Again, keep asking about how can we get more R&D. I think, again, we adjust all the packages, our salary, our benefits, our care about our employee. Okay. Headquarter in Zhunan, we keep have expansion. New building ready by end of this year, able to increase 1,500 space for R&D. In Taipei office, we also keep expansion from 300 to 500 headcounts.
We are going to also establish our Tainan office, already start going to operation, already got 30 headcount there. By five years, the goal is to get 500 engineers in Tainan and also the Taipei, because a lot of science park people prefer to work in Taipei. We also establish our Hsinchu office, R&D office, by this month. We try also to recruit more R&D from third-world country. Okay. Keep working with the university. We sponsor the professor and the lab to have a co-development project. Also we are increasing our R&D in the U.S., Japan, and also China for local design-in operations. Investor also asking, Phison has a very good success and also story in the Gen4. How about Gen5? We repeat, we reuse our model, our strength on the Gen5.
We already work with the two semicon and also with the CPU partners to enable our Gen5 SSD platform, not controller, but also with the modules. By next year, 2022, when the CPU platform ready, Phison also ready with the Gen5 client SSD. On the other hand, we're also going to have our Gen5 data center controllers co-work with our very closest partner, try to get the business in the data center by Gen5. Gen4 silicon is ready, system ready by end of this year. Gen5 controller will be ready by next year. Hopefully 2023, Phison will be the key player in the data center storage. One thing, half years ago, we announced Phison is going to participate into the Gen5 Redriver and Retimer. We start to have R&D development on ReDriver by eight months ago.
After eight months, we got the samples two weeks ago. We test everything. Looks like performance is over my expectations. First cut, bug free, one of the best performance right now according to my motherboard customers. We are going to directly go to the MP immediately. After this week, we share our report to the key player in Taiwan. We already got the designing projects, two projects ongoing, and start to talk about shipment next year. I feel happy, proud of my people. First cut, first design five months, bug free, directly to the MP. Now we have a confidence in the Redriver business, and next year, second half, we will have a Retimer ready to the market. This will become our new business.
Again, revenue-wise, this is not able to contribute high revenue, but this will help Phison to have a more influence power into this computing systems. Imagine, Phison, we have a Gen4 SSD, Gen5 SSD, data center SSD. If we are able to make the most service, Gen5 redriver retimer means Phison, we are able to play very good position in this ecosystem. Also CPU makers start to recognize Phison not only in the storage but also in the high-speed interface solutions. This will help Phison to gain more business in coming years. I believe by next year, redriver, Phison will be the key player. 2023, Retimer Gen5, Phison will become the key players. Again, a lot of investors concerned about the inventory, if we need to increase our revenue, our top line, we need to have manage on inventories.
When you look to the inventory level, right, revenue keep growing, inventory definitely we need to follow. Because of the whole supply chain out of control today, components in the past, lead time four weeks, eight weeks, now go to the 14 weeks, 15 weeks. In order to make sure customers able to get the products, we need to increase our inventory to make sure we are not facing the shortage. We need to build up the inventory. I still believe my inventory today is still healthy. The reason is, I wish I able to grow our revenue through our well controlled, well managed inventory. The last about the ESG. For many years, Phison, we didn't recognize we needed to pay attention on the ESG. We didn't do the good job, and especially in the MSCI, their rating.
We've been CCC for many, many years. Since two years ago, recognize this is important, we start to invest, start to focus, build a committee. According to the OTC's ranking, Phison is the top 6-20 by 2020, and we are upgrade. By the market value over TWD 10 billion in Taiwan, we are also ranking from 60%-20%. By MSCI now we're ranking to BB. I believe by end of this year, we're able to get the other upgrade. We focus to make the more R&D investment to get the more business, increase our influence, power in the market to get the more profit return to shareholders, employee. Also, we try to invest and focus more into the ESG management. This is all my presentation today. We can go to the online.
Thanks, K. S.. In the meantime, if you have any questions, please use the ask question text box and I will help to read. If you feel okay, I can also share your company name. Actually, I have some questions to ask K. S. first. I'm very interested about one of your slide about the R&D intensity. There are kind of a good way or kind of a negative way to read this, right? Because your R&D intensity is like 60%. It's much higher than your industry peers. Supposedly you should acquire a much better market share or much higher ASP, given the more investment in R&D. It doesn't seems to be the case today. I know company has that potential, but can you please justify that kind of very high R&D intensity in a consecutive two years? Thank you.
If you ask me today, my R&D intensity today, I'm hard to answer you by this tracer record. By 2016, 2017, you can see the trend. R&D intensity we are increasing. You look to our 2027 new business, the performance. Looks like we are out-performed. Means we have to invest more in R&D, we're able to get return after two to three years. I believe with our hard work in the R&D, we get the today we prove we make it. With today's our R&D intensity, by 2023, personally, I believe we're able to get the more influence, power, business and revenue. Again, I always say to my people, the goal is TWD 100 billion, and we need to make more money to make sure we are good in the investment, R&D investment. Why Phison keeps investing that many resources?
Our silicon partners, our system partners, industrial partners, every partners keep asking Phison, they need the new products. Unfortunately, I talk to my R&D managers, they say, "Hey, not enough. Overloading. Our people is going to crazy." What can we do? We have to invest. We have to hire. Of course, if I do it today, the cost increase immediately. If I'm not doing today, two years later, I won't get the new project, new business. Again, you trace to the Phison's our business model. Phison's able to make a good profit through the business model, not only controllers, but also modules and the systems. We have a good profit, we invest this into the R&D to make sure after two years, Phison will be able to make more profit to this company, to the shareholders.
Again, I'm hard to answer my investment here, what is going to happen. You look to today's our results. Two years ago, what we did, then we got the return. Again, don't worry, we know how to manage. Yeah.
Okay. Thanks. It seems like you achieved the headcount targets earlier, right? 2,500 engineers. Are you going to change your targets, meaning even more than 2,500 people in the coming five years?
Actually, from early this year, Taiwan are facing the type of the talent and TSMC saying that they need to get the extra 6,000 engineers. The population here, the young student here, the number is there. We don't believe in early this year, we're able to get 400 people. We don't believe, but just half years, seven months, we make it. I think we make it because Phison at least we have a good reputation in Taiwan. Our result performance, I think the public, they're able to recognize. Our strength and our future space and storage, a lot of young talent looks like they agree, they buy in. We hit 400 in half years. I believe the other 400 by second half. By 2022, mid-2022, definitely we've hit the 2,500.
Again, by next year, I still believe we may request by our partners, silicon partner to get the more design input for them. That's why we prepare our CB. We got the other new land. We plan the other new building in two years, and this is to prepare for the other growth in the headcount. I believe the answer is yes. By 2023, we may hit to 3,000. Okay? Not I'm driving the headcount. I drive by our partners to push me to get more people to service them.
Okay. That's great to know. There are several questions from several analysts. Maybe we start with CLSA, Brian. He's asking, how does the PCIe redriver retimer complement your exceeding solution? Can you share the TAM? Where you are going to expand your interface business to other interface, for example, USB and HDMI. Thanks.
Why Phison, we jump into this redriver retimer business. Actually, again, we are starting with Gen5, not Gen4, not Gen3. The reason we decided to go to this business, just because Phison, all the IP, high-speed interface IP are doing in-house. We just reuse our IPs and reuse our know-how. Actually, when engineer proposed to me, they want to build redriver. First, I asked, "What is a redriver?" I didn't know what is so-called redriver. They said redriver is a needed in the coming Gen5. Every motherboard, server, PC, they need redriver. I said, okay. They want to invest. I said, "Okay, you do it." They say, "They have a confidence." I said, "Don't tell me confidence. Show me the products." They show me products after the product by mid-July. Two weeks.
Well, I can say it's perfect. I lead these two meetings with the motherboard PC makers. They also surprised. Now I got confidence to my people. To be honest, I don't know what is the TAM of redriver. I don't know what is retimer, I just know all the new Gen5 system, they need both redriver and retimer. Every. Okay? Phison, definitely we are going to participate all the Gen5 storage, no matter in the server, PC, notebook, automotive, whatever. I don't expect, seriously, I don't expect redriver helping Phison to get the big revenue. I believe there was some certain revenue and a good profit, this is not my first goal. My first goal is we need to increase our influence power into our customers.
They need storage, we're able to provide storage, but we have extra offer, extra service, high-speed interface solutions. With this kind of ecosystem supply to them, I believe more system partners, they need Phison. This is also helping Phison in the future to get more design win projects into the system, which is able to cover high-speed interface and also the storage. This is what I'm doing. Again, I have totally no idea how big the market, who is competitors, but I know who is the customers. Okay, who can become. Every PC maker, server maker are the customers. We simple. We just send a message, talk to them, just have a call. We present to them one hour, they're surprised. Their eye turn big. When they see our test report, they ask for samples.
By the way, actually, we already got two design win for next year. Two design win. Start, first cut, I see just before two weeks now, we are discussed how to ramp up the production by CQ1 next year. Anyway, good signal. We get a good starting, so I have more confidence in the retimer. Just side story, my business development head count, people of this project in my Phison, he proposed me to buy a report from some consulting company in a few $10,000. He need to study the report. I turn down. I said, "Not necessary. Why need report? We just go to talk to the system makers, computing makers. They can tell you what they need." We just go to their way.
Maybe after six months, I can tell you how big a TAM, what is the competitors, how to do it, but right now, I only concern how to bring this first cut products into the system makers. Again, already two customers are surprised, so I feel good. Okay.
Okay. Thanks, K. S.. Next question comes from, I think, should be Bank of America, Mr. Simon Woo. He's asking about where you see foundry sourcing issue from TSMC, UMC, or SMIC, given the current shortage. Which foundry do you have a bigger exposure, you say that UMC or TSMC? Maybe we start with these two questions. Thank you.
Okay. Phison, we are only working with the UMC and TSMC. Now we also have a partner with GlobalFoundries. Okay. Is our new supplier. Overall, we keep very close relationship with the UMC and TSMC. They give us a good support. The wafer amount gradually increasing. Unfortunately, the demand is too strong. Far less than enough. I don't know how to convince them. I keep telling our foundry partner, Phison has very unique position in the market. Okay. Now you see, we have the other new redriver retimer to service customers. Means customer in the future, they need more from Phison. Phison, we also willing to invest the money into R&D to develop a lot of new system, new products.
Based on this kind of a position, I believe Foundry people, they may agree to support us more because we are willing to invest, we are willing to expand the business, and we are trying to build more products to the industry. At the end, they are going to benefit. I believe we are able to get increased supply, but unfortunately, demand is still too far than enough. We will keep doing our best at this moment.
Okay. The follow-up question from Simon is about your PCIe chip geometry. You said 14-16 nm nodes that were still at 28 -30 nm range. Thanks.
Gen3 mainly in the 28 nm Gen4, now we migrate to the 12 nm. Looks like all the design is starting from 12 nm. We also have IP design today, with the 7 nm, Gen5.
Oh, okay.
Yeah.
Okay, got you. We have several new questions that I maybe speed up a little bit. There's a question from Brian at CLSA. What do you expect the ASP profile by segment in second half 2021? I guess it means the ASP increase by segments. Do you think that the SSD controller IC shortage will continue through 2023? If so, do you expect the ASP trend to go up in 2022? That's the question. Thank you.
Phison, we have a different portfolio here. So you're talking about ASP. Consumer is the other story, okay? Actually, a controller embedded system, gaming module, industrial, the ASP are slightly increasing because of cost increase and material shortage. Overall, Phison, our business except the consumer, ASP increasing. You're talking about controllers, ASP, basically, Phison, our controllers, actually Phison cover all kind of controller from the USB, eMMC, UFS, memory card, and the SATA, PCIe Gen3, PCIe Gen4, we cover full range. Overall, every range we are in tight supply. We are shortage. This cover mainly, 55, 40, and 20 nm. Y ou may not know, Phison also have in-house PMIC solutions. All the SSD supplied by Phison controllers and modules are using Phison in-house PMIC. PMIC actually also in tight supply. We're not enough.
There's mainly in the 8 in, 118 nm. Next year, I still believe 8 in, 55/40, definitely tight. 28 nm, I think slightly tight. Phison is more far than behind because we got so many design wins. Mainly in the eMMC, we are key players. Foundry, we keep talking to foundry, try to get more capacity. By 12 nm, I think also under constraint. Phison is mainly Gen4, is picking up so rapidly. This, we also need our customers, our partners, helping us to talk to the foundry to get the supply. ASP-wise, I believe next year because controllers are still tight, so ASP able to maintain. I believe. Yeah.
Hey, thanks K. S. . The next question comes from Randy Abrams from Credit Suisse. He is asking how much capacity increase have you secured for next year, with the LTA? How much higher price would you commit to, I guess, to your foundry customers? Can you also secure your business with the LTAs with your customers? This is the first question from Randy.
Okay. Yeah, we have two plus one foundry partners. Sorry, I'm not able to share it to you how many percent I'm able to increase, this is too competitive, and I don't want to bring trouble to my foundry. Everyone go to talk to them to increase the capacity. For sure, Phison, our supply next year is increasing. I believe still keep gradually increasing. Again, still far behind. That's why we have to work very close with them to convince them based on our position, they need to help us to get the capacity. Talking about LTA, actually, most of my key customers proposed to make the prepayment to Phison, and they asked to have LTA with Phison to secure the controllers. This is a two side. We talk to suppliers for LTA.
Our customer also talks to us about the LTA and prepayment. We need to fully use our resource capacity to support our key partners. Once we make the design win, we need to fulfill the demand. What is challenging? Demand is too strong. Actually, I wish the semicon market has some downturn. We are able to get some released capacity from others. I wish. Yeah.
Thanks, K. S. . Second part of Randy's question is about how is your view on NAND? Do you expect a normal slow season in the fourth quarter? Do you think the NAND market can stay tighter with the industry strength? Maybe I also moderate a little bit, if you can qualify the NAND price in 3Q and the first quarter in both the raw NAND and module price. That'd be very helpful. Thanks.
The NAND supply already tight for over, I think it's four quarters since the last Q3. Personally, I don't like to see NAND always in tight supply because this is not healthy to the industry. NAND industry, actually, the human keep asking for more capacity. If the supply is tight means the price is not able to go down, means the market is not able to grow, especially the 5G start to penetrate to the life today. More storage is needed into their systems. This quarter, basically, NAND still in the strong demand, tight supply because of data center looks like start to picking up. Smartphone by Q2 was slow but looks like now still back to normal. PC slightly start getting soft, again, the demand still there.
Next year, 2022, according to the supplier saying NAND still in tight, demand still good. Again, if the whole year keep the NAND tight supply is not really good to the industry. I wish you can see that some more new fab is going to ramp up. Some supply is going to come in. In the future, the NAND price by annually cost down 40% + in the past. Okay. We wish the price able to, by gigabyte, cost down to attract more system to use a more high-capacity business. This is much healthy. Overall, NAND is strong. Price-wise, I think a lot of report. By Q3, the component itself, supply increase 5%-10%, which I do not see this. Anyway, supply is tight. Yeah.
Okay. Thanks, K. S. . We still have four questions, and we have eight minutes left. See if we can answer all of them. If not, I will pass the question to the IR team. The next question coming from Howard Wong from Glory Capital. How will the gross margin and OP margin evolve in the next few years when we ship out our efforts in the new products?
You can see Phison. Phison has IP, data system business. Don't say Phison gross margin is 50%, at our last call, our goal is 25%+, luckily, the previous two quarters, we hit 30%. In Q3, I am still optimistic we are able to keep 30% something. Phison, we need to increase influence power. We have to increase our timeline in the revenue. My goal is, again, the margin is a 25%+. The best is 28%-30%. This is our goal. At the end, the earnings per share, we need to increase a lot. This is what we have to do. I am not much afraid about my gross margin rate because I am able to generate more gross profit through our business model.
Okay, thanks. Next question come from Wedbush Securities, Mr. Matt Bryson. His question is, roughly how much of Phison's business comes from industrial and embedded? It seems like those are business is going through a period of acceleration. Can you comment about the business potential there?
I think we have a portfolio here, right? Embedded now is 14% of total revenue, industrial 19%. We can see this field, automotive, is going to use more high-capacity products. Revenue-wise, we can expect. Gaming modules, I didn't mention, is going to terabyte. Embedded, mainly we get into the PC, computing, server, and also the smartphone. This, actually, if Phison going to spin off either one unit, this can be one independent company and still able to grow. This is already a big segment in the industrial. Just because Phison today is still young and starting, give us few more years. I think we're able to grow this few segments by ourselves. Again, Phison not going to spin off. We need to keep everything into Phison with our unique business model.
Okay. I will skip one supply-demand of the NAND flash question, and jump to the follow-up question from Mr. Simon Woo. First of all, if you look at the control IC market, can you comment about the market share this year versus last year? Also, you talked about the controller IC price hike over 20% QoQ in second quarter. What's the trend into third quarter? Those are the key questions.
Okay. Again, this is difficult from Phison's side to tell you the market share, because I read the other company saying they have a market share, but we go to collecting the total amount, looks like this is too far than what they can supply. We do know how many pieces are in-house from Samsung, from Hynix, from others, right? We got no numbers, I'm not able to tell you the number. By unit wise, we are increasing. By the supply, we're increasing, but again, unfortunately, still far and behind to ship to customers. If foundry able to give us 2x of wafer, we still able to ship immediately to customers, means the demand is very strong. Phison, what we more care is about PCIe Gen4 for investment. We are in a good position, 2022, 2023.
This, we are going to make the more controllers for our semiconductor partners. On the other hand, we're also able to build the modules to increase our gross profit. Again, don't get me wrong, I'm not talking about gross profits percentage, I'm talking about gross profit. We able to get the more dollars profit from the modules, industrial modules, gaming modules, embedded modules. When you look to Phison, right? That's why our finance is strong, because, again, of our business model. Controller-wise is our key business. We need the more wafer. On the other hand, we enable the controller to semicon. We also enable the modules without any extra cost. That's why Phison is able to create more profit.
Okay, thanks. Second part of Simon's question is about the module ASP. He asked very specifically, was your second quarter module ASP up 7% or 15% QoQ? For third quarter revenue, do you expect quarter-over-quarter revenue can grow around 10% quarter-on-quarter?
Phison, we cover full range of modules. We are not only provide SSD or eMMC, we cover everything. It's difficult to me to answer how many percent increase average ASP. The answer is, ASP is increasing because of controller constraint. Okay. Not only because of price increase, but because of a content capacity increase. To be honest, Phison, we don't like to see the same gigabyte storage price keep increasing. This is not healthy. Okay. We hope we able to slightly constant quarter-by-quarter the modules, but we need to increase the content. Means every 18 months, we're able to double the content to the customers.
Okay.
Overall, we increase the ASP, but the dollar per gigabyte are declining. This is the most healthy in the industry.
Okay. Lastly, I summarize the two questions from William Sung and Matt Bryson into a simplified question. What do you think about the NAND demand supply in 2022? The IC or foundry demand supply in 2022.
Foundry, no need to answer. It's tight, super tight. I wish it can be getting loose, but I am able to get more supply. NAND, okay, according to Semicon, they said very strong. This is what they said. From my point of view, I still believe next year they have some new fab coming. Overall, the COVID still is the uncertainty, and the price increase too much in the NAND may also decrease the demand in the content. It's hard to say. Anyway, next year, the NAND still in a good shape, good position, but price-wise, again, I don't expect the price keep going up. Okay? Not healthy.
Okay. Thanks. Yeah, maybe we need to finish here. I'm sorry if I didn't moderate your question very well. I will make sure I send your questions to the company management, and I hope they can answer you by email. Thanks, K. S., and also, thanks for your all the participation. Have a good day. Thanks.
Thanks, Charlie. Thanks to everyone. Thank you.
Bye-bye.
Bye-bye.