Hello, everyone. Welcome to Alchip Technologies 2019 Institutional Investor Meeting. This meeting will be in English. If you need the Mandarin version, the Chinese version of the slide, please go to the MOPS, the [Non-English content], to download the presentation material. I wish all the participants would please write down their questions to the host through Zoom's message function. Please write down your name and your company's name with your question. We will answer five questions verbally. If you want to ask questions verbally in the Q&A session, please use the raise hand function through the Zoom application, and then I will unmute you for your questions. The video and the audio content of this meeting will be uploaded to MOPS [Non-English content] right after this meeting when we finish the recording of the video.
Because of the coronavirus, we are so sorry that you will miss this important Institutional Investor Conference. I think for the first session, I will hand it to our CEO, Johnny Shen, to do a brief opening for our company and some 2019 results report. Thank you for participating in this meeting. Okay, Johnny, please.
Right. Okay. Okay. Ladies and gentlemen, thanks for coming to the investor conference meeting. My name is Johnny Shen, CEO, and Co-founder of Alchip Technologies. It's the very first time for us to host the conference meeting online. Due to the foreign investment percentage increase, we decided to use English to run this meeting. Honestly, companies still prefer to have a face-to-face meeting instead of online. During this special virus situation, please bear with us, and thanks for your understanding. A quick go through our company's current status. I believe most people are already familiar with Alchip. The company was founded in 2003. We went public in 2014. Currently, we have 450 people, and the company is still growing. By the end of this year, we are shooting for 500 or plus.
75% of our employees are engineers, so we are a purely engineer-oriented company. Since we founded this company in 2003, we've successfully taped out more than 420 designs. Last year's number, we have achieved TWD 140 million. We also have more than TWD 22 million in R&D investment. We are one of the TSMC agents; we call ourselves Value Chain Aggregators. Next slide, please. I believe most people should understand the ASIC market now. The ASIC market is very big and unique. Unlike a standard product company, one chip needs to serve multiple customers. By definition, one ASIC chip only serves one product. Outstanding product companies need to make their own chips in order to improve the product efficiency and also differentiate from their competitors. Alchip has been doing the ASIC business since we founded the company. We focus on the HPC and AI area.
HPC and AI always require the most leading-edge technology. Currently using 16 nanometers or seven nanometers. Eventually, we'll use five nanometers or below. The chip itself is very big and complicated. The entry barrier for our competitor will be getting higher and higher. We truly believe the big data accumulation is too fast and also the internet bandwidth improves. Industry always needs a more powerful and efficient chip in order to analyze the data. This application will continually grow and be sustained. We also expect the profit margin will be reasonable since most of the HPC and AI companies are highly profitable. I believe this application will continuously bring reasonable profit to the company. Okay, next slide, please. Let me recap the achievement for the last year.
Basically, we have a conference meeting every quarter and also a smaller group meeting hosted by Daniel, sometimes weekly, sometimes biweekly. At the end, we meet our revenue target. The revenue number is TWD 140 million. We're up 22% year-over-year. Net income: TWD 14 million, EPS TWD 7.2. Both of them are record highs for the company; they are historical highs for the company. In capital market performance, we are also doing a very good job. Share price appreciates more than 3x, also with very high liquidity. Our investor weight for the foreign investment is getting higher and higher. Also, institutional investors keep increasing. Company achieve all quarter, even monthly profitable. Last year, as you remember, we started to penetrate the U.S. region. We also have very good progress in the U.S. We've been successfully penetrating all target AI users.
Right now, we can proudly say every major AI user knows about Alchip. For sure, we will be one of their solutions, one of their choices, if they want to do the AI chip. We gradually diversify the dependency of the China region. In fact, we already won a few key accounts, even with the five nanometer opportunities. Next. We have a great position in the HPC and AI fields. Right now, the majority of our revenue, so-called more than 70%, is contributed by the HPC and AI. We have six designs in high -volume production right now. Multiple seven designs, a seven nanometer design ongoing. That will gradually taper off from this year. In fact, other than those ongoing designs, we have many designs in the pipeline for seven nanometers.
In addition to the seven nanometer design, we also have many 22, 16, and 12 nanometer design wins, and some of them are under bidding. Last recap and also the highlight for the company is our crisis management and control. We have a limited impact but huge benefit from the trade war. The Chinese government and city continuously inject a lot of funding into the CPU and HPC. Most of our customers are beneficiaries of the China investment, and we see more and more orders from them. Great reaction during the coronavirus situation. Government just announced an additional two to three weeks of holiday. Our engineer, 90% of our engineers, could work from home efficiently. As you know, our database is all sitting in the cloud. It makes no difference for our engineer working from the office or from home since all the databases are sitting in one place.
At the end, all projects are secure with no delay. That's a quick recap and also a review for 2019. Right now I will pass the detailed numbers discussion to our CFO and speaker. Daniel, please.
Okay. Welcome, everyone. For the financial session, I will state our last year's performance. For the numbers, I believe for you, since we already announced every month's top and bottom line, the revenue and the net income should be well-known by the market. I still brief a little bit. For the fourth quarter last year, the total revenue is TWD 47 million, which is a 61% quarter-on-quarter and 83% year-on-year growth. For the operating income is TWD 4.6 million for the fourth quarter, and that is 35% quarter-on-quarter growth. For the net income, since we have done government subsidies that came in last four quarters, the net income is TWD 4.6 million, up 24% quarter-on-quarter. Translating into EPS of TWD 2.34. The year 2018 is a good year for us.
The total revenue is TWD 140.1 million, up 22% year-on-year. The net income is TWD 14 million, up 64% year-on-year, translating into EPS of TWD 7.2. The following is the breakdown for the application and the technology node and the regional breakdown for Alchip last year. For the quarterly breakdown, you can see it on the left-hand side. This time we separate the HPC from other applications. It's a new category for our application breakdown since we didn't reveal it in the previous conference meeting. Everyone can see that the HPC revenue accounts for a majority of our total revenue for each quarter last year, especially for the third quarter and the fourth quarter. For the whole year, the revenue comes from HPC, accounting for 59% of our total revenue, followed by networking, 14%, and the niche market, 18%.
The consumer now only accounts for 9% of our total revenue in 2019. For the AI application, the AI application is within the HPC field. The 59% include the AI and the CPU. For the process node breakdown, you can see the quarterly breakdown, it's very obvious that the seven nanometer and the 16 nanometer have already become the majority of our revenue source. For the advanced technology node, you can also see the 28 nanometer or more advanced technology node already accounts for 90% of our total revenue in 2019. For the seven nanometer, in 2019, it accounted for 27% of our total revenue. I believe such performance should be leading within our industry. For the revenue breakdown by region, it's pretty straightforward. You can see the Korean, now we don't have too much business from Korea.
Obviously, because of China 's CPUs and the other projects, the China region has already become the number one revenue contributor to Alchip. For the others, you may see the others keep on growing for the past four quarters and for the past three years. The reason behind it is the U.S. market, we have been very, I would say, successful in U.S. market penetration. I believe the revenue contribution of the U.S. market will gradually surpass the contribution from Japan, maybe one or two years later. For the industry view, many investors focus on our China CPU business and keep on asking how the coronavirus will affect the shipment. Far from it, I would say the customers' current forecast remains unchanged. Even though the coronavirus epidemic and the geopolitical change between China and the U.S. For now, the forecast is still very strong until the year -end.
If we look into the quarterly pattern, the first quarter will be higher, and the second quarter will be down a little bit. In the third quarter, the China CPU shipment will recover to maybe above a one -quarter level. The fourth quarter will still remain strong. In addition to the current shipments of two to three chips to our China customers, the next -generation projects are kicking in. Actually, we are already doing the design for three , four, and seven -nanometer projects for our China customers. For the AI market demand, which is still strong. Although, for now, there are no significant production shipments currently. The new design demands are getting stronger and stronger for the North American market , for the Middle East, and for the China region.
The shipment for AI chips, I believe, will gradually kick in starting in, I would say, the early second quarter this year. The shipment to the North American customers and the shipment to the Middle East customers will begin in the second quarter this year. As the other projects enter the take-off phase, we will see an uptrend for the AI shipments this year. We have key accounts even with the five-nanometer opportunity for design. For the business outlook, I think China CPUs; I think everybody already acknowledged that the China CPU shipment is still the major revenue driver for this year's revenue growth. First of all, the shipment will be four-year shipments.
Last year, we shipped the China CPU in only the first quarter, which was kind of sluggish, and a very small amount for the second quarter last year, and a small amount for the third quarter last year, and a pretty good amount for the fourth quarter last year. For this year, every quarter will be good for the China CPU shipment. Of course, the order size is expanding. We expect the China CPU shipment to drive this year's revenue growth. For this year, the shipment for the seven-nanometer CPU chips could be the critical factor in the second half revenue growth. Based on our schedule, we will start to ship out the seven-nanometer CPU type chips to our customers starting in the third quarter this year.
Since the seven-nanometer capacity of TSMC is very tight right now, we cannot make sure so far about the capacity support from TSMC. Since it's getting better and better, we are currently optimistic toward the support from TSMC, since the overall consumer market demand for seven-nanometer capacity is weakening. Honestly, at the current stage, we are becoming more optimistic. This year, the seven-nanometer projects will become our mainstream technology nodes. Actually, this kind of breakdown already began starting from the third quarter last year. You can check it from the process node breakdown in the previous slide. There are multiple seven-nanometer projects from China, we are either already tape -out or we are in the design phase. We have multiple seven-nanometer projects from Middle East customers. All of them are AI type customers. We also have multiple seven-nanometer projects from North American customers.
For the tape-outs, the seven-nanometer projects, based on our schedule, we also believe there are multiple seven-nanometer projects. I would say more than three to four will tape out this year. According to the information, we are currently very optimistic towards our revenue growth this year. I think our CEO can do some outlook -closing statements from now on. Johnny.
Okay. Yeah. Just a quick recap. We reserve more time for the questions. Basically, 2019 was a great year for the company. Yeah, in fact, it was the best year so far. Based on our current outlook, if we count out all the uncertainty, for example, the TSMC capacity issue or any unavoidable disaster, we have confidence that 2020 is going to be another record-breaking year for both the top line and bottom line, for both production and service. Right now, just like Daniel said, seven nanometers have become mainstream in our product line. We have a lot of current projects. The current projects will ensure our revenue grows for this year. The strong pipeline projects will keep the momentum going all the way to next year. Overall, I think 2020, again, is going to be another wonderful year for the company. All right. Thank you.
Okay, now is the Q&A session. I haven't seen the written questions. I only see the hand-raising. Let's start from the verbal one. [audio distortion] Okay, Charlie, you are unmuted. Please.
Okay. Thanks for taking my question. Can you hear me okay?
Sure.
Okay, thanks. Thanks for hosting this webcast. I think it's quite efficient, and congratulations for your good results and execution. First of all, can you give us some more quantified revenue and margin guidance for the first quarter and the full year? Also, I'm also very curious about user feedback for those local PC/CPU systems. Do you think that demand is sustainable? Can we start from here? Thanks.
Okay. For the revenue and the target and the margin guidance, I would say , Okay, I think many people already heard me from the meeting in our company that our goal for this year, the revenue target, is TWD 200 million in revenue for the top line . For the gross margin, it's hard to tell right now because the percentage of NRE has shown great influence on our gross margin. So far, based on our estimation, we think the margin will not be as high as the level of last year, because last year, our NRE percentage was almost like 60%, close to 60%. For this year, in the beginning, I would say the breakdown for NRE and production, the line will be like 40% something for NRE and 50% something for mass production.
The blended gross margin will be lower than last year's level. For now, the current production margins look a little bit better than we previously thought. I think our gross margin level will be around the mid-30%s. For the user experience, honestly, I didn't use the PC that was embedded with our customer's CPU. I will hand this question over to Johnny to see if Johnny has any input. Johnny, please.
I got it. Thanks, Charlie. Overall, I think, as you can see, there's a revenue difference between Q4 and Q3. Last year, in Q4, we had more significant growth. I believe this momentum will last to Q1 as well. I think the momentum will remain. Overall, just like Daniel says, we are shooting for a break of TWD 200 million for this year. HPC and AI mass production will kick in. Like I mentioned before, the profit margin for these two fields, even the mass production profit margin for these two fields, is much higher than consumer. Also, we have more and more design NRE, more and more design opportunity. Based on the current estimation, just like Daniel said, 45, 55 for the service and mass production, but we do see some momentum for the additional service revenue. That will happen later this year.
Q1, I have confidence; I think Q1 is going to be a very good quarter.
Okay, thanks. On that local PC system, actually, besides the end -user feedback performance, it would be very helpful if you could give us some color about the total addressable market in China, who the end customers are, and the potential volume this year and the next year. That would be very helpful. Thanks.
Daniel, would you want to take this?
Okay. Charlie, honestly, I don't have that kind of information for you because we don't know the user experience. As I know, because the Phytium CPU is the PC embedded with the Phytium CPU is installed with the Kylin OS. Currently, Kylin OS is the most friendly operating system for ARM-based PCs or notebooks. I guess if we compare the user experience between ARM-based and Windows-based, I would say, of course, the Windows-based will be better. Within the ARM, I believe the Phytium CPU plus Kylin OS will still be the leading user experience provider for the ARM-based notebook market. For the addressable market, so far, the answer is still the same: that we think the scope now is the SOE company. If every SOE or part of the SOE, we don't know.
We believe the SOEs with critical information or related to national security, for example, the financial sector and the telecom sector, or maybe some sector with critical information, will accelerate the adoption rate for China CPUs inside PCs.
Okay. Thanks. Yeah. That's all for my question. I will be back to the queue. Thank you.
All right. Thank you very much.
Mm-hmm.
Jeffrey, I will unmute you. Can you hear me, Jeffrey's iPad two. Excuse me, Jeffrey, J.I., can you hear me? Okay. No. Okay. Since there's no response, I will answer some questions from the written ones. Okay. Many people are still very interested in our revenue pattern. I would say for this year, unlike last year, we have a solid production revenue contribution throughout the whole year. The revenue pattern will be the first quarter, I will still expect a flat or a little bit of an increase. Quarter-on-quarter increase for the first quarter. The revenue, compared to the first quarter last year, still leads me to believe there will be growth. For the second quarter and third quarter, it's relatively flat.
In general, we may see mild quarter-on-quarter growth for the first quarter, second quarter, and third quarter this year. We expect a relatively significant jump for the revenue for the fourth quarter since the seven-nanometer production is scheduled to start shipment in the late third quarter of this year. That's the question many people are asking about. The second question is on the operating expense plan for this year. For last year, our operating expenses were about TWD 37 million to TWD 38 million. For this year, since we are still investing our R&D resources, the design, engineering, hiring, and especially the machine. For a seven -nanometer or even a five -nanometer project, it consumes a lot of machine power.
Because of that, I personally expect the operating expense for this year will be ranging from TWD 44 something million to TWD 46 something million, depending on the recruiting and the machine usage. Okay, that's the second question. Also, people are asking about the seven -nanometer project profit margin. I think I will split into the NRE and the production. For the design NRE, the absolute amount of profit is the highest among all the process nodes. For the gross margin, since for seven nanometers the mask is expensive, the IP is also expensive. For the gross margin, the seven nanometer actually is lower than maybe the 16 nanometer or the 28 nanometer for the gross margin rate, the percentage. For our company, I guess many of you know already that we capitalize our mask and the IP.
Most of the time, when NRE revenue kicks in, it's almost 90%-100% gross margin, depending on the milestone. The mask and IP will go to our fixed cost, not the amortization cost. I believe this answers your questions very well. Okay. Chiho Wong is asking about the CapEx guidance in 2020. Actually, if you look at our financial report, the majority, over 90% of our CapEx, is the mask and the IP. As long as our business is doing well, the CapEx will increase. It has a very high correlation with our NRE revenue. Matt is asking about the first quarter this year, the NRE and the mass production breakdown. The revenue breakdown for the first quarter, I think, will be around 40 something for NRE and 50 something for MP.
It really depends on the scheduling of our NRE milestones. Here I would like to remind everyone that for the seven nanometer project, a single milestone is easily TWD 2 million to TWD 3 million. If that's a tape-out milestone, it's easily TWD 8 million or TWD 9 million. If any of the milestones are delayed or preponed, it will affect the revenue and especially the margin a lot for the single quarter or for the single month. Yeah. That's our view of the first quarter. Okay. I guess Jeffrey is back. I will unmute Jeffrey.
Okay. Jeffrey, can you hear me? You are ready to speak. Everybody can hear you right now. Hello? Jeffrey from Macquarie. Jeff? Okay. Technical difficulties. I will answer the question directly. Jeffrey is asking us to discuss more on the networking -related business. Honestly, we have not put too much focus or attention on the networking right now. I know 5G is the hot topic for the market. We don't think the 5G market is a supermarket for ASICs. I still believe the 5G market belongs to the ASSP product. Currently we have two projects related to 5G. One I can remember is the Piconet, it's like an extension networking kind of chip for 5G networks. To us, this is not a big project. We don't expect significant or very quick growth from this sector.
Let me add some comments on the 5G. Like Daniel mentioned before, we are not focused on the 5G client application, but we do have a few customers interested in the cloud, in the station side. As you know, the 5G bandwidth increased a lot, but the distance becomes shorter. All of our customers consider, even on the station side, the volume will be higher than 4G. I'm looking forward to seeing some contribution starting from later this year on the 5G station. That's the comment. For the Q&A, I think you're welcome to raise a question either in English or Mandarin.
Yeah, we can understand Mandarin. No problem.
Okay.
Okay. Here is a question. I think Johnny can add some comments on it. Investor asking about the Middle East and North America AI market. Can you give us some sources about the customers and the timing for production and the future outlook for these two markets, the Middle East and North America?
Okay. Yeah. In the Middle East, we are doing a good job, as you know. Currently we are under an NDA. We cannot disclose the customer's name, I think a few famous customers in Israel are our customers. Their company is also doing very well. Actually, they have a great exit already. Currently, two 16 nanometers are ready for production. In fact, their seven-nanometer design is ongoing and will be tape-out very soon. Yeah, we have no problem continuing the service to their mother company. In fact, the new seven-nanometer design will be tape-out soon. Another big design will be coming, the contract is already signed. In the Middle East area, in addition to this customer, we have two to three additional customers.
U.S. I think we already won a few. The few major players for the cloud service are our customers, and their product is ready for production. Yeah, in fact, they already have small-volume production right now. Based on their forecast, we are looking forward to having a very reasonable amount of production starting from the middle of this year. They will start to make a contribution starting from the second half. In parallel, their next generation is ongoing. We will tape out before the end of this year. Other accounts, we are also penetrating. All the giant companies, like I mentioned before, know about Alchip. They have a lot of pipeline. We are looking forward to seeking a big contribution from the U.S. region. Yeah, they will surpass Japan soon, and sooner they will be the second highest region, right behind China.
Okay. I think this question is also to Johnny. That in our view, is there any chance that the Chinese government will accelerate its demand for the local CPUs going forward?
In fact, we've been working with our customer very closely. To be honest, our customers are very optimistic for their market. As you know, for mass production, they need to place the order three months earlier, so we can easily tell on Q1 or Q2. Q3 and Q4, right now, we don't have a clear picture, but based on the discussion between us, they have confidence, but all the conversation will go back to prior to the Chinese New Year. Right now, I think all of China is suffering about the coronavirus prevention. I hope this one will not influence their demand. Overall, for the first half, we already received a steady order from this customer, and the momentum is very strong.
Okay. Here is a question asking about a seven-nanometer CPU. The one that is scheduled to go into mass production in the later quarter this year. The question is asking, is this CPU for a server or a PC? I can answer you, for the CPU planning to start mass production late this year should be within the server area, not the PC NB CPU. For the PC and the NB CPU, the CPU, currently, we're shipping our products to our China customer, Phytium. This product life cycle will be longer than we expected. According to the current plan, the next -generation product will be scheduled to tape out maybe in the second half next year. We can only see the production starting maybe near the very year-end of next year, 2021. The current generation's shimmer will last at least two years.
Okay. If there are any questions, please use the raise hand function of Zoom, or you can send me a message. Far, I guess we already covered the questions investors are asking. Again, sorry for the inconvenience because of the coronavirus. Okay. Okay, people are asking. I guess this question should belong to Johnny, that an investor is asking, seven nanometer is tight. Seven nanocapacity, I think, is tight. How do you get the capacity from TSMC?
Oh, that's a very good question. Okay. Seven-nanometer capacity is very tight, everybody knows. Right now, this year, both the CPU market and also the HPC, majority of our designs in production are 16. I see the insignificant impact on the production revenue on the first half of this year. Second half, we already start to communicate with the TSMC on seven-nanometer capacity. In the past, if we followed the TSMC game rule to give them a six -month prior forecast ahead, they never failed us. Seven nanometers, there is an uncertainty. Yeah, we can only continuously work with the TSMC and based on the relationship and also past experience, we should be able to get most of the forecast out. Yeah, because all our customers are very big, the major AI users and also the CPU users.
Customer and we will work together continuously and also frequently meet with TSMC in order to ensure the capacity.
Okay. Another question that I can answer. Some investor is asking about how many customers account for more than 10% of our total revenue in China. There are two customers. One, I guess everybody knows, is Facebook. Another one, sorry, we cannot reveal the names. We have a very strict NDA. There are two customers. Each accounts for more than 10% of our total revenue in China. There is a hand-raised question. Oliver Li, your name on the Zoom is Oliver Li. I'll unmute you.
Thank you very much. I'm Oliver from BlackRock. Can you hear me?
Yes. We can. I will hear.
Okay. Hi. You just mentioned the seven nanometers actually have lower gross margins than maybe the corporate average. Just wondering, are the reasons for lowering margins because of the higher CapEx tools expense or additional R&D? Maybe the scale is not sufficient to amortize the expense, so maybe over time, the margin will get back to the corporate average or even higher. Is that possible? Does it happen that way?
Okay. Oliver, I may need to correct my answer ; when we talk about the margin, I assume the mask and the IP are the costs. To our book, as I mentioned, we capitalize mask and IP. For the milestone, when the milestone happens, we recognize the milestone revenue. Most of the time, for the milestone revenue, we have almost a 90%-100% gross margin. The land, the mask, and the IP, since they are capitalized, will go to amortization cost, which is our fixed cost. Usually, we amortize our mask and IP for no longer than 18 months. Why is my answer the seven-nanometer margin is lower, because many people are asking about the margin for different technology nodes. I can only say, under normal situations, our NRE, our engineering fee contribution, accounts for, let's say it this way.
Our engineering efforts, the money we charge, accounts for a smaller portion of the total contract value of seven nanometers. The absolute amount of the value of the effort is bigger than, say, 16 nanometers or 28 nanometers. Can you understand me?
Yeah, understood. Let me clarify this. After you amortize the IP and tools, and then if we just look at the financial statement, the margin probably looks similar because of the amortization. Is that right?
Sorry. Can you repeat?
No. You just mentioned if you just treat the IP spending as an expense, then the margin might be lower in that way.
Margin rate.
Sorry, the margin rate. Yeah, margin rate might be lower. After amortization, the margin rate might be similar to—
It will be higher.
It will be higher.
Right. Oliver, let me explain it this way. Yeah, the seven nanometer, of course, the design efforts are higher, and the entry barriers are higher. In terms of design fee and also production margin, it definitely will be higher than 16 nanometers.
Yes.
The reason Daniel mentioned it is because there's a huge cost for the mask and IP. Yeah, we all know the mask for seven nanometers is much, much higher than 16.
Majority of our mask fee, we need to pay TSMC. That mask fee, we have a very limited margin. For our design and production, if we count out the mask fee, the profit margin and design margin are much higher than 16%.
I see. Understood.
Yeah. If the production lasts longer, just like you say, past 18 months, past all the amortization period, all the production volume becomes very high. I expect the margin will also be higher.
Okay. Yeah. That's clear. Thank you very much.
Okay.
Okay. I don't see other questions written or hands raised. If there are no questions, thank you to everyone for joining our institutional investment meeting today. Again, sorry for the inconvenience of using this online format for the presentation. As long as the coronavirus epidemic is ongoing, when the situation gets better, we still prefer face-to-face conference meetings for the future. Thank you, everyone. I will stop the connection. For the whole video and the audio, again, we will upload it to the MOPS. If you are interested, you can watch the recording online. Thank you.
Okay, thank you very much. The next meeting, we are looking forward to having a face-to-face meeting as usual. Yeah, in fact, we really miss the face-to-face private conversation and communication by the end of this meeting, but this time it's not going to happen. Yeah, we won't be able to exchange business cards as well. Anyway, thank you very much. Thank you.