Semiconductor Manufacturing International Corp

HKEX:981

Earnings Call Transcript 2018-Q1

[Operator]

Welcome to Semiconductor Manufacturing International Corporation's First Quarter 2018 Webcast Conference Call. Today's conference call is hosted by Dr. Haijun Zhao, Co-Chief Executive Officer; Dr. Mong Song Liang, Co-Chief Executive Officer; Dr. Yonggang Gao, Chief Financial Officer; and Mr. Tim Kuo, Director of Investor Relations. Today's webcast conference call will be simultaneously streamed through the Internet at SMIC's website. [Operator Instructions] The earnings press release is available for download at www.smics.com. Webcast playback will also be available approximately 1 hour after the event.

Without further ado, I would like to introduce to you Mr. Tim Kuo, Director of Investor Relations, for the cautionary statement.

[Tim Kuo, executive]

Good morning, and good evening. Welcome to SMIC's First Quarter 2018 Earnings Webcast Conference Call. Today, our CFO, Dr. Gao, will highlight our financial performance and give guidance for the next quarter. And then our Co-CEOs, Dr. Zhao and Dr. Liang, will provide some business commentary. This will be followed by our Q&A session. As usual, our call will be approximately 60 minutes in length. The earnings press release and financial presentation are available for you to download at www.smics.com under Investor Relations in the Events & Presentations section.

Let me also remind you that the presentation we'll be making today includes forward-looking statements. These statements and other comments are not guarantees of future performance, but represent the company's estimates and are subject to risk and uncertainty. Our actual results may differ significantly from those projected or suggested in any forward-looking statements. For a more complete discussion of the risks and uncertainties that could impact our figure -- future operating results and financial condition, please see our filings and submissions with the U.S. Securities and Exchange Commission and The Hong Kong Stock Exchange Limited, including our annual report on Form 20-F filed with the United States Securities and Exchange Commission on April 27, 2018.

During the call, we will make reference to financial measures that do not conform to Generally Accepted Accounting Principles, GAAP. These measures may be calculated differently than similar non-GAAP data presented by other companies. Please refer to the tables in our press release for a reconciliation of GAAP to the non-GAAP numbers we will be discussing. Please note that all currency figures are in U.S. dollars, unless otherwise stated.

I will now hand the call to our CFO, Dr. Gao, for financial highlights.

[Yonggang Gao, executive]

Thank you, Tim. Greetings to all our listeners. First, I will summarize our first quarter results, and then give the second quarter 2018 guidance. In the first quarter 2018, our revenue was $831 million, an increase of 5.6% quarter-over-quarter, mainly due to the recognition of $108 million in technology licensing revenue. Our revenue in the first quarter 2018, excluding the technique -- technology license revenue, was $723 million, a decrease of 8.1% quarter-over-quarter, mainly due to the product mix change, lower ASP and a decrease in wafer shipment.

Gross margin was 26.5%. And if excluding the effect of license revenue in first quarter '18, the gross margin was 15.6%. Non-GAAP operating expenses were $196 million. Profits for the period attributable to SMIC was $29 million, while noncontrolling interest was $2 million offset to SMIC's attributable profit. EBITDA was a record high of $325 million.

Moving to the balance sheet. At the end of the first quarter of 2018, cash on hand, including other financial assets, were $2.3 billion. Gross debt-to-equity was 50%, and net debt-to-equity was 16%.

Now looking ahead into second quarter of 2018. Our revenue is guided to be up 7% to 9% quarter-over-quarter, including the forecast to recognize the technology licensing revenue estimated at $56 million. If excluding the effect of license revenue in Q1 and Q2, revenue in second quarter is guided to increase mid-teens percent quarter-over-quarter. Gross margin is expected to range from 23% to 25%. If excluding effects of license revenue, it's expected to be 18% to 20%. Non-GAAP operating expenses are expected to range from $227 million to $233 million. Noncontrolling interests of our majority-owned subsidiaries are expected to range from positive $17 million to positive $19 million, which are losses to be borne by noncontrolling interests.

The planned 2018 CapEx for foundry operations is adjusted from $1.9 billion to $2.3 billion. The increase of CapEx is mainly for R&D equipment and the expansion of capacity of 8-inch fabs in Tianjin and Shenzhen. The planned 2018 CapEx for nonfoundry operations is adjusted from $48 million to $137 million, mainly for the purchase of land and the construction of employee's living quarters and our new headquarter. Our planned 2018 D&A is approximately $1.1 billion, mid-teen increase comparing to previous year.

I will now hand the call over to our Co-CEO, Haijun, for general remarks.

[Haijun Zhao, executive]

Thank you, Yonggang. Thank you all for joining us on today's call. SMIC is undergoing a period of transition. As mentioned in the last quarter, we are confronting many challenges. However, through the efforts of recent quarters, I'm pleased to say that things are looking better than originally expected with the demand picking up, utilizations rebounding and an encouraging progress on R&D and the business platform development.

In first quarter, our revenue grow by 5.6% quarter-over-quarter and 4.8% year-over-year, mainly due to the technology license revenue of $107.6 million from the Shaoxing project. The Shaoxing project is a startup company located in Shaoxing, Zhejiang Province, in which SMIC has joined [indiscernible] state as a minority shareholder and a JV partner. In line with our strategy to focus our resources on key platforms, some of our specialty technologies have been licensed to, and will be utilized by, the Shaoxing JV project. When excluding revenue from technology licensing, our revenue decreased by 8.1% quarter-over-quarter, in line with industry seasonality. This decline was mainly due to the seasonal weakness, particularly in the smartphone sector. Although there was a general decline in output, we still benefited from the sequential growth from power IC, RF, connectivity and the NAND Flash-related devices. Gross margin, excluding technology licensing, was at 15.6%, higher than originally expected as utilizations made a turn for the better.

Our revenue in first quarter from China region grew by 28% sequentially and a 40% year-over-year, and when excluding the technology license revenue, grew by 2% sequentially and 11% year-over-year. We firmly believe that our position as the foundry of choice in China continues to bring us great opportunities.

As artificial intelligence, electrical vehicles, autonomous driving and IoT become more prevalent, China will be a focal point of considerable activities. We forecast that the China fabless market will continue to grow around 20% per year for the next couple of years. We are well positioned to capture the meaningful prospect and can expand our site -- addressable market opportunities by accelerating the development of our technology.

Given our progress in recent quarters, we have increased confidence in our technology development execution, which include both advanced nodes and mature nodes platforms. Both our power and image sensor based in these platforms are among the leading in the foundry industry, providing us our customers with competitive services and technologies. In addition, our Flash memory business, which has utilized a strategic customer model, is now one of the key revenue drivers for this year. Our revenue from power, image sensor and the Flash memory grew more than 30% year-over-year in the first quarter 2018.

We work to develop our business platform into a comprehensive services offerings in areas that aligned with meaningful opportunities stemming from the China markets. Our current view on this year is co-business is better than when we last spoke on the fourth quarter earnings call. We see positive signals from current orders, and our customers have forecasts yet remain cautious. We continue to target revenue growth of high single-digit percentage. Apart from technology license revenue, growth drivers, including power, NAND Flash and the connectivity-related ICs, will maintain our profitability targets of annual teen growth margin and annual profitability attributable to shareholders. We are happy to see our technology development has been progressing well. With the accelerated activities and the improved business outlook, we have raised our CapEx guidance to USD 2.3 billion from the original guidance of $1.9 billion. This increase is mainly for the manufacturing equipment, R&D equipment and the facility constructions.

In closing, we are in a period of transition, but we are optimistic with utilizations bottom up in the first quarter and a general momentum picking up in orders. SMIC is moving quickly to align with customers to capture the opportunities before us. We take each step with vigor and caution and strive to advance SMIC to the next level, while targeting profitability.

I'd now turn the call over to our Co-CEO, Mong Song, for further comments.

[Mong Song Liang, executive]

Thank you, Haijun, and thank you to everyone on the call for joining us.

Over the last few quarters, we review our business markets, organizations, resources and capabilities, and we have formulated an overall strategy which we had begun to implement through changes in the company's day-to-day work culture. Our long-term strategy and focus are to build a solid foundation for SMIC starting with a strong organization and culture, which stretch for continuous and innovative improvements, our aim is to instill a driven team culture.

Meanwhile, we accelerate the development of our technology and aim to build up complete technology platforms, which integrate competitive technology, ready-to-use IP and comprehensive design service in order to increase competitiveness and capture timing of meaningful market opportunities.

I am pleased to say that I'm increasingly confident in our R&D team's execution as we swiftly meet each of our key milestones. Today, I'm going to highlight the progress of our technology development and platforms in more detail.

First, our 28-nanometer versions are progressing well as we work our way one version at a time. Our 28-nanometer HKC, which entered the production in the second half of last year, has seen tremendous and rapid improvements recently, reaching industry competitive years. We have quickly reached our 28-nanometer HKC+ R&D milestones, and we aim to start production in the second half of this year. Instead, we continue to enhance our 28-nanometer offering to provide better power and performance. 28-nanometer continues to be an important node for the industry, which is mainly market size. 28-nanometer is forecasted to continue to grow as an increasing number of applications begin to migrate into the 28-nanometer node. As the largest foundry in China, we are in a position to benefit from these coming opportunities.

FinFET technology will also be an important area of growth in the semiconductor industry in the coming years. We accelerate our technology development, striving to close the gap between us and our customers' advanced node requirements. We first target to serve our customers' needs for FinFET, springing from the low-end mobile and digital consumer markets. We are pleased that we are pulling targets for FinFET development, as mentioned last quarter, in our target risk production for our first version of FinFET to begin in the first half of next year. Our aim is to provide customer with easy node migration, various device integration and full IP coverage. Our mature technology platforms, we have strong position in power IC. I'm pleased with our team's progress on power IC, which has met notable improvements in increasing solvency and reducing mass grades. We are making continuous enhancement to our BCD technology to provide competitive services and aim to develop a 300-millimeter BCD offering.

With our mature platform development, we strive to be a leading foundry source with diversified applications and customers. As we evolve technology portfolio, we continue to enlarge our addressable market size.

Finally, to conclude my remarks, to date, we are working harder to accelerate, execute and deliver for the future growth and profitable development of the company. We believe the current investments will translate into business as we aim to gain share and expense get. Thank you to all our listeners for your continued support, and looking forward to giving you our future updates. I will now hand the call back to Tim for the Q&A session of this call.

[Tim Kuo, executive]

Thank you, Dr. Liang. Today's Q&A will be hosted by our Co-CEOs, Dr. Zhao and Dr. Liang, and our CFO, Dr. Gao. I would now like to open up the call for Q&A. [Operator Instructions] Operator, please assist.

[Operator]

[Operator Instructions] Your first question comes from line of Randy Abrams from Crédit Suisse.

[Randy Abrams, analyst]

First question is 2 parts. For the second quarter where you're guiding pretty good strength or rebound up mid-teens, could you talk about if that's being driven or how much is being driven by 28-nanometer rebounding from the first quarter correction? And how much coming from the mature applications? And which applications there? And the second part is for the full year outlook. I think you maintained the full year outlook, which ex licensing would be low to mid-single digit growth, but curious with the second quarter pickup, what your view is rest of the year, why you're staying with the original outlook.

[Yonggang Gao, executive]

Randy, thank you for the questions. For the first question, for 28-nanometer recovery, you know the first quarter is the traditional seasonal season for the mobile phones. And our 28-nanometer PolySiON dominated by the mobile phone products. So we saw the corrections of inventory in the first quarter. And second quarter, we see it's better than the first quarter. And innovation, we've already been running our 28 High-K Metal Gate products for a while. We continue to remark the time delay, mainly because the long lead time of 20 High-K Metal Gate machines from our [ wonder ] side, but we're already build up the capacity in a certain part and continue to build up the 20 High-K Metal Gate capacity. So in the second quarter, we make a full use of our existing capacity for 28, both PolySiON and the High-K Metal Gate. So that's -- you see that we already forecast, there'll be a recovery there. And -- but you know for the existing capacity, the ASPs are pretty big from the competitive market, so we do not mean that we will recover to the first quarter level, but we really go for high single digits. And for your second question, yes, second question, for the second quarter, we forecast a better outlook and the revenue come from where. Actually from all the technology platforms SMIC already build up, and I already mentioned that 28-nanometer and the 40-nanometer, 55, we also saw SMIC's NAND Flash, MCU, CMOS imager, and we also mentioned the RF or Bluetooth, more or less we say that's because of the recovery of the overall market, especially the connectivity markets. And the demand for the power -- analog power, very strong. I can say is possibly because the electrical cars and others, they need more capacity dedicated to decrease on the power devices and IGBT, but making the whole industry, 8-inch, and the power devices capacity is already in tight situation. And then I believe everybody in the foundry markets benefit from the over-demand of the power devices. And then the utilization go for almost full loading for everyone.

[Randy Abrams, analyst]

Okay. And the second question, just the last part of it, I guess you maintained the full year outlook, but it sounds like there is good strength on 8-inch and then a decent rebound on 28. So I guess for the full year, if there's any view, why not stay as it has more visibility in the second half? And then the second question was more on the increase in CapEx. If you could give an update how that translates to capacity additions, if now -- it sounds like it's more tied to 8-inch, but if you can give an update, like which fabs would be expanding and how much capacity you had this year.

[Yonggang Gao, executive]

You know that we are expanding better demand of our strategic customers, and currently, we have very strong demands for 8-inch capacities running from 90-nanometer copper all the way to 0.25-micron, 0.35-micron power devices aluminum technologies. And part of our -- on the CapEx, we'll move to Shenzhen fab. We'll build up Shenzhen fab to full capacity. We also complete the clean rooms starting up in Tianjin new wafer fab where we equip our fab with additional capacity to analog, power. And at this moment, we sold -- because of the delay of the equipment shipments and the expansion in 8-inch, we will try our best to equip these 2 fabs. And our foundry fab has been running the commercial CMOS imager technologies well and the logic. It has been running on full capacity there. We try to move up capacity to better meet customer's forecast. For 12-inch, you know that we are running pretty well, 55-nanometer connectivity things, and we are expanding the NAND Flash, NOR Flash, CIS and our power devices in that area, in relation to just now I mentioned the 28-nanometer High-K Metal Gate capacity. That's these things -- so that is the capacity expansion you asked for. Yes, that means we expand 28 High-K Metal Gate capacity and our 40-nanometer and 55-nanometer, 65-nanometer to meet these special requirements sold in devices I just mentioned. For the 8-inch, we build up our fabs in Shenzhen, the 8-inch fab. I would also like to make a very big round up in Tianjin new wafer fab 8-inch.

[Randy Abrams, analyst]

Okay. And I guess your full year outlook, you didn't change it, but it sounds like there's better strength. But could you talk kind of what you're seeing right at this stage for second half? If it's more of a pickup in second quarter but keeping conservative on the incremental ramp in second half.

[Yonggang Gao, executive]

Okay. And actually, you know a lot of things happened on the market. Even though we have a lot of preparation in the diversification of the technology platform, I'll also mention, we are quite mainly focused on analog, power, PMIC, fingerprint and CMOS imager, Flash, high V, [indiscernible] High-K Metal Gate. And we do believe that we're well prepared for the market change, but because of the additional things in the market, you know that -- and also because of the delay of the equipment shipment for our additional capacity and also the new capacities coming up from the competitors, we do not have very good visibility into the second half, but we still obtain mistake and very cautiously aiming at high single percentage annual growth.

[Operator]

Your next question comes from the line of Charlie Chan of Morgan Stanley.

[Charlie Chan, analyst]

So my first question is regarding the recent issue between U.S. and China, that trade war. So for example, if the U.S. will set a ban on ZTE or Huawei, what would that mean to SMIC in the mid-term and long term? Is this company is likely benefit from more domestic semiconductor replacement? Can you give some comments on that?

[Haijun Zhao, executive]

Charlie, thank you for the questions. So first things first, SMIC has been a very international company. We follow strictly the commitment we already made to and the technology and equipment vendors to SMIC. So we should say, for SMIC, we are in a very good situation, and we're very -- and are confident in that. And you know that SMIC does not produce any products. And this kind of limitation on the applications on the productions, mainly happens to our customers' side. And SMIC has been maintaining a well balanced between international customer, IoT customers and the domestic customers for the past many years. You can see our revenue segmentations that we maintain a very big chunk of international customers just to balance, maintain a balance of that. And our customer in China are very, very diversified. So we see that the impacts from this round of trade tensions are very limited to SMIC. And for our guidance to second quarter, we already include this type of impacts. And with the trade going, I believe that we can slowly minimize the impacts and the effects of the impacts.

[Charlie Chan, analyst]

Okay. Okay. And my next question is about your comments on 28-nanometer, pricing pressure and also the power, IGBT opportunity. So on that 28-nanometer, can I interpret it that there will be some -- more ASP pressure? I guess your key competitor, TSMC, also said they need to build up their fab, right. So I guess that's where the ASP pressure comes from. But in that case, you seem to have bigger revenue scale, lower ASP. What does that mean to your gross margin trend in 28-nanometer? Do you think that gross margin trend is worse than your -- or better than your previous expectation?

[Yonggang Gao, executive]

For 28-nanometer technologies, you know that our customers do need SMIC's [ specialty ] technology and continue to grow the capacity and the performance in this area. So we're already committed to our customer. We will continue to run 28 High-K Metal Gate and the PolySiON and the further technology developments there. And yes, the competition in the markets of 28-nanometers is very, very tense. And we saw the ASP erosions and the pricing pressure, but we already made that commitment. We will continue to do this production. And to cope with the pricing pressure, on the one hand side, we will do better in technology performance, give the added values. And on another side, we would do the flexibility in the production launch to make sure that we will run the 28-nanometer technology and the capacity can run -- can share with the other technology, like a 40-nanometer, and we can share with a larger manufacturing base and lower the cost.

[Charlie Chan, analyst]

Okay. So in that case, what will be your new gross margin guidance for full year then?

[Yonggang Gao, executive]

Yes, just now, I already said that gross margin guidance, we will go forward, and yes, it's a teens percentage level.

[Charlie Chan, analyst]

Okay, okay. IGBT, what's your progress here? Do you already got the IDM outsourcing with a significant revenue contribution? Or power, IGBT is just a new opportunity you're looking for?

[Yonggang Gao, executive]

Actually for the power devices, I mean, decreased power devices and IGBT type of markets are showing a very, very strong demands. We believe these markets will continue growing. The power management for the artificial intelligence, high-performance computing for the motor controls, for the automobiles' charge station, et cetera. And we also sell the requirements from the customers' side. But for SMIC, for past many years, it is difficult for us to run power devices, IGBT in the same fab with the other devices like logic, CMOS imager, RF. So for SMIC, you know that we already announced the building up of joint ventures outside of Shanghai. And they will set up their facilities and to ramp up these kinds of specific technologies.

[Operator]

The next question comes from the line of Leping Huang of CICC.

[Leping Huang, analyst]

I'll ask in Chinese and then follow [Foreign Language] So my first question is about the 28-nanometer. So what will -- we see the revenue of 28 dropped quite significantly this quarter. So what's the reason? And the way you recover from second quarter, what will be the move there, application and the major mix between the Poly and the High-K Metal Gate?

[Haijun Zhao, executive]

Thank you for the question. And for the first one, just now, we've mentioned that for 28-nanometer in the first quarter, actually at the beginning, the fourth quarter time frame, we worked very well with our customers to forecast the first quarter, second quarter 28-nanometer demand. And the happenings in the first quarter, and especially in your fab rate, and we see -- with all the adjustments on the inventory. So even though we're already starting a lot of wafers, but customer hold that, hold them and hold the shipments, these kind of things. So the final results in the first quarter is the 28 revenue just go down. And with the recovery of mobile phone markets and ramping up the 28 High-K Metal Gate, we really see the demands of the second quarter for the shipments. So we already forecast that we will go for high single-digit of shipments in second quarter. And the demand for 28 High-K Metal Gate at this moment that we are running mainly consumer loading of products, and the demand is high. We are very short at -- in the third quarter, 28 High-K Metal Gate revenue and the shipments that we're housing the 28 PolySiON.

[Leping Huang, analyst]

[Foreign Language] The second question is about the SMIC's business strategy in memory market. So what -- can you elaborate the detail about your innovative or the business model on these current NOR Flash? And what's your future plan with it? Do you have interest to enter other bigger markets like NAND or specialty memories?

[Yonggang Gao, executive]

Leping, you know that -- first things first, the business model. You know that nobody can really draw a logic foundry for memory foundries. And in the past many years, quite many people tried, and then they just gave up. And the strategic type of our business model just refers that we need to form a strategic alliance with our customers and targeting either specific market area. So that we can invest the money and the resources to do the technology development and to a high-quality level. So currently, for SMIC, we do not do anything commodity or generate products with customers. And what we did and we work together with customer targeting specific niche market and with high qualities, especially for servers and automotive. And this applied to both NOR Flash and NAND Flash. And just now you mentioned the NAND Flash is a small market. That's very true. For specialty NAND Flash, the market is relatively small. And without further growth possibly, they'll be stable there. And the things that -- not that many players does either. So for the established players there, they do have the market shares there. And as far as we can provide a very high-quality performance there and we can continue the base mix. And the things is fairly same for specialty NAND Flash. And the demands actually grow very fast for specialty NAND Flash. And we also use a similar model strategic alliance with customer targeting and high-quality and niche markets and develop a specific technology for that. And we've already been running 30-nanometer NAND Flash specialty and a 24-nanometer specialty for quite a while and running very high quality and a good year.

[Operator]

The next question is from Steven Pelayo of HSBC.

[Steven Pelayo, analyst]

Let's see, first question, I guess you talked a little bit about targeting profitability, but I'd like to try to explore, what's the breakeven for the model right now? In the first quarter, if we exclude the licensing revenue and the $17 million in R&D credit, it looks like kind of an ongoing operating losses basis, operating profit would've been a loss of about $60 million, $70 million or so. And it looks like, I guess, from the second quarter, if I exclude the licensing agreement again and with the higher OpEx and even the 20% gross margin, we're still probably losing money at the op level. So can you talk a little bit about what we do see in terms of revenue and gross margin to generate a breakeven at the operating profit line? Once again, excluding things like R&D, subsidies and extra licensing revenue.

[Yonggang Gao, executive]

Steven, that's a very good question actually. That's the formula and the calculation every day I'm doing. And 3 factors: the first factor is the ASP versus the quality and the performance and the competitiveness on the markets. And some are very hot nodes that many people are working on there, and the ASP is changing every day, and there's a lot of factors. When we plan, the debt factor is very hard, and we intend to remark, and this is the first thing. The second thing is the portfolio of the equipment or depreciation of the hot ASICs. And if we invest a lot of new ASIC and they get into heavy depreciations and definitely the breakeven ratio will guide in higher. But if we lower down the CapEx spending and lower down the numerous [indiscernible], maintain for a while, we can guide in better utilization and lower depreciation, and the ratio can guide it lower -- I mean, lower ratio, guide at breakeven point. And currently for SMIC, we set out our targets for gross margins, for NAND profitabilities and also it comes into customer demands and a forecast for the nice -- 1.5 years to 3 years. And with this kind of factor fit in, we control our pace in procurements of new machines. We calculate -- more or less, we go for the allocation type of [indiscernible] and profitability. We come up with the amounts and control our spending on a pace. For example, if I give the forecast high teens of gross margin, and we have the ASP erosion, we calculate every quarter for the next few years and we have a lot of communication with our customers and also the depreciation, many things. And when we have the new request for capacity spending and we have new project to set our new wafer fab getting new capacity, we're putting this kind of thing into the number to see what's the results. If the results go for noncontrollable, and we have to make sure that either we can have a mirror to get a cost reduction or we slow down or delay the new spending. So I share with you the things that factors to ASP, the quality of SMIC performing and the depreciation and the expansion plan we have. So with this kind of dynamic factors in our hands and we should [indiscernible]. In the past couple of years, we do have a lot of the spending on capacity building up, the depreciation guiding higher and with this kind of depreciation we calculate, we guide the number for the gross margin. And once we committed the gross margin to the market, we will maintain this, too. That means cost reduction, ASPs and the place for further spending. Yes, that's my answer to you. That's the dynamic...

[Steven Pelayo, analyst]

Actually, I was looking if you can quantify a bit more...

[Yonggang Gao, executive]

You can calculate the number out from our gross margin, these kind of things. Utilization, we also announced and gross margin we also announced. You can know right away, today, today, was our breakeven point. Last year was our breakeven point. And with our non-GAAP spending and our commitment to the numbers of gross margin, you can calculate what's the next breakeven point.

[Steven Pelayo, analyst]

Okay, why don't we just refocus a little bit on 28-nanometer then? I was really confused, were you -- did you say that you expected high single-digit growth quarter-on-quarter in the second quarter? Or you expected 28-nanometer to be high single digits as a percentage of revenue? What were you suggesting for the second quarter?

[Yonggang Gao, executive]

The second, the second. That means our 28-nanometer shipment revenue will recover to high single digits of total. And we do not -- I did not really calculate how much percentage it has paid us off. I just now mentioned high single digits, the revenue ratio, yes.

[Steven Pelayo, analyst]

Okay. So it was roughly 3% of the first quarter revenues, and now you're saying it's going to rebound to high single digits. I guess what does that mean then for the remainder of SMIC? Is it above 28-nanometer, actually flat to down for you? I haven't done the math here, but what does that suggest then for above 28-nanometer?

[Yonggang Gao, executive]

We should say this way, we continue to build up 28-nanometer High-K Metal Gate type of capacity because our customer requests that we need to make sure to fulfill our commitment to them to support their business. And what we are doing today to make the flexibility to make a full use of the buildup capacity. So when the markets really get frustration seasonally, and we still have the other applications just now I mentioned, 24-nanometer, NAND Flash, and the other applications. Because we build our 28 capacity, they can share with the 40-nanometer. 40-nanometer, we have even more applications like MCU, embedded Flash, CMOS imager, NAND Flash and NOR Flash and high Vs. So we can always use other fillers to fully use the capacity we newly build up. But the expansion mainly depends on -- because 28-nanometer, we already have the technology customer there, mainly depends on the business. We do not have a subjective number to say how big we need to build up. We mainly build up to the stage that customer currently work with SMIC. We follow the markets.

[Steven Pelayo, analyst]

Okay. Do you have an estimate -- last question from me, just -- when you look at the full year 2018, if you're guiding to revenues to grow kind of high single digits quarter -- year-on-year, how much do you think 28-nanometer gross year-on-year or as a percentage of revenues for the full year? And then once again, then it begs the question, what are the revenues, excluding 28-nanometer growth? So can you talk a little bit about full year outlook for 28-nanometer alone and then for above 28-nanometer?

[Yonggang Gao, executive]

At this moment, just now we said, we follow the market demands and fulfill our commitment to our customers, and we recover to high single digits and possibly we maintain, possibly, I say possibly, and we'll maintain this kind of high single digit through the year. And by the way, we're really open to the additional demand from our customers.

[Operator]

Your next question comes from the line of JunJie Chen from Tianfeng Securities.

[JunJie Chen, analyst]

I'm from Tianfeng Securities, and I actually have 2 questions for Haijun and Mong Song. And my first question is for the mature node. On the 12 -- for the first year of the mature node on the 12-inch, I noticed -- do you see any product change from 8 inches to the mature node of 12-inch because I see the Texas Instruments, they use 12 inches to do some products -- for analog products, so that they're cost-effective. But we -- I noticed that you mentioned that you have the capacity and CapEx for the 8-inch but you do not have any CapEx spend compared to -- on the mature node of the 12-inch, [indiscernible] or technique [indiscernible] for the product transition from 8-inch to the mature node on the 12-inch. [Foreign Language]

[Yonggang Gao, executive]

Okay. Actually, you mentioned a very good point. 8-inch type of wafer fab, the total capacity will be limited. You know that nobody really build up brand-new 8-inch. So in order to -- we have 2 challenges. On the one hand side, we need to make sure that the 8-inch keep competitive. Another thing that we have to meet our customer requirement for additional requirements on capacity that are currently running in 8-inch. We haven't found a solution to that. And we are trying. And you mentioned that like the industry leaders, already use 12-inch to do the analog in [indiscernible], that's entirely true. And the [indiscernible] many customers choose to use 8-inch, mainly because, on the one hand side, there is [indiscernible], especially in design, testing, packaging, everything is on 8-inch. The second thing is that 12-inch, at that moment, we're still very expensive. 8-inch are fully depreciated, and 12-inch are still running high depreciation. But today, the situation changed. 8-inch can now meet these requirements and the 12-inch also finished the depreciation cycles. And we started -- we already worked together with our customers to run -- first things first, to run exactly same technologies between 8-inch and 12-inch, such as fingerprints, CMOS imager, analog, power, BCD, etc. SMIC has been successful in running the same, exactly same products in both 8-inch and 12-inch. And the only concern is 12-inch whether or not we have enough capacity to take in additional orders to cover 8-inch. The second thing is whether or not a customer can have the supply chain in the back-end for bounding, bunking, for stacking and for casings and because 12-inch back-end capacity probably were also very expensive and very limited, and 8-inch back-end capacity were overbuilt. They were very cheap. And if we can find the justifications and they will do -- for example the back-end bunking kind of account for 15% cost reduction, if we are running both 8-inch and the 12-inch. If they can find a good solution in the back-end customer, they will want to go for 12-inch. So to answer your question, in general, yes, that's the translation, and SMIC has been doing that. And I believe that in the future stage, we'll continue to run both that way. And the new Shenzhen, we're running 12-inch together with 8-inch wafer fab. And in Tianjin and in Beijing, actually for the same product, we're running both Beijing 12-inch and the Tianjin 8-inch, and the business model and technology transfer has been proving successful.

[JunJie Chen, analyst]

Okay. Okay. My second question is for the 40-nanometer. And I just want to know that how to balance the revenue and the process for the 40-nanometer. Because I see the 28-nanometer, the contribution of the revenue is okay, but the profit may not contribute too much. And when we do the -- if we do the 40-nanometer, our competitors, for example TSMC, may also leverage downside of their ASPs, and we may also be facing the same with the investor demand -- for example, TSMC has installed a wafer fab in Nanjing and the 60-nanometer capacity is ready. And so I think if we have the risk production in the first half in 2019, and we're also facing the ASP problem and how to balance the growth rates and the revenue. [Foreign Language]

[Haijun Zhao, executive]

Thank you for the question from Mr. Chen. This indeed is a very difficult question, and I've been thinking this for quite a long time, okay. At this moment, of course, our 40-nanometer is not in production yet. So we're talking about the revenue and profit is a little bit early. So my thinking is that in 2 phase, okay. First thing is, how do we learn from the 28-nanometer, High-K Metal Gate development lessons not to repeat that development time, too long and missing the better market window and resulting at this moment the pricing pressure? So our thinking is, therefore, new technology ramp up, normally we need to have about 3 phases, okay. First phase, definitely our cost is higher than ASP. And second phase, that we will try to balance, make ASP and costs crossover. And the third phase, we definitely will aim for ASP higher than the cost. So in order to do that, we have to carefully adjust our production volume. For example, first phase, we can assume it's running about 5k per month; second phase, about 10k per month; third phase, about 15k per month, in that ballpark, okay. So in order to make that ASP and cost and cost crossover, we have to carefully select our market segment, our products. And of course, we also need to watch our intrinsic cost and also enhance our performance, reduce our power to attract customer. So in terms of the product segment, we're thinking, for example, in the 40-nanometer, at first phase, we will try to focus on the high-end consumer and media applications. And for the second phase, we're thinking to go for the mid-, low-end mobile application, and we also work, prepare for the AI and some ASIC IP for cryptocurrency, to mining application, to blockchain application. But for the final surface, in order to make the higher ASP, we will develop the RF applications. So that's just an example of our production ramp up and our product segment selections. And all of these we are discussing closely with our customers. So we understand we shouldn't repeat 28 High-K Metal Gate current situations, and we have to, in execution phase, in fact, we have restructured our R&D organization and arranged more resources and just focus on the essential technology. So we believe the 40-nanometer will bring a new phase to the SMIC in the near future. And that's the answer to your questions.

[Operator]

Your next question comes from line of Bill Lu of UBS.

[William Lu, analyst]

I'm hoping to get a little bit more clarity on the guidance. First of all, you talked about 28 nanometers going from 3% to high single digits. Can you give us the same math for power and also for Flash? What was it as a percentage of sales in 1Q? And what's your guidance for 2Q?

[Yonggang Gao, executive]

Bill, in general, we do not go for so much detail. Actually, I do not have the data in front of me to give more detail for the different platforms, the contribution to growth or revenue. Instead, we gave the overall guidance for overall revenues. But I can assure you that the analog, power devices platform, we have multiple platforms ready for production also developed further to add value to these kind of platforms. And for the Flash, as I already mentioned, that for the NAND Flash and we're mainly targeting further. And for the high quality, especially for automotive and solar applications and make it a steady production and maintain the ASPs there. And for 24-nanometer NAND Flash, we will remark further. And the growth will be there, but I do not have the data or breakdown to say what's exactly the growth for the next couple of quarters. But we do mention probably delay that we sold for the Flash over -- year-over-year bonus 30% growth in revenue.

[William Lu, analyst]

Okay. I guess I'm just going to think through the gross margin trends. If you look at this year, it seems like as 28, do you think it should go up, margins should start to improve, and I've got to assume that power margin is above your corporate average. So why shouldn't margins go a little bit higher in the second half of the year? Is it depreciation? Or can you help me with that?

[Yonggang Gao, executive]

The factors mainly come from 2 sides. The first side is from the depreciation. You mentioned that for the couple of years, we have been spending and in the CapEx in building the capacity. And this kind of depreciation, we strictly follow the financial accounting rules that really started from the show up. That's the burden of building up new wafer fab, building up new capacities. And the second thing comes from the utilization. That means the first quarter, the wafer orders, the total demand for SMIC's capacity lower than before, probably we're almost running 100% loading, full loading. But first quarter, our utilization is low. And for some technology nodes, actually the demands just shape from one node to another. And we need time to ramp up to fulfill the capacity, I mean, the new capacity and the capacity they gave up to new applications, and I guess, mainly come from these 2 things. And the computation on pricing, the price erosion also play a [indiscernible]. And for this part, SMIC has been trying every year for the cost reduction to compensate and to balance, cancel out this kind of a price erosion. That play a little bit better, but know the big portion, the big portion number one, comes from the depreciations and the number two come from the overall SMIC wafer fabs utilizations, and we already say that third quarter utilization is not that good.

[William Lu, analyst]

Second question is on China fabless. You talked about a 20% CAGR over the next several years. Now from the outside looking in, it seems like a lot of the new growth is coming from crypto and AI. I'm sure you got a much better view than I do. I'm wondering if you agree with that, or if you could help me a little bit with the growth areas. And then if you look at where the fabless industry is growing in China versus your own capabilities, it seems like a lot of the crypto AI guys want leading edge. How do you think about your own development next several years versus what the customers are going?

[Yonggang Gao, executive]

We should say this way, the growth of China fabless comes from all fronts, all labels. And if you look at it today, [indiscernible] the IGBT, [indiscernible] in mobile phones have set up these kind of consumers and from all fronts, and the ASIC high-performance compute, like you've mentioned, would be common things. It's pretty new. We also learned that some system companies like Alibaba, like Baidu, like Tencent, they're also thinking about building up their capability in designing ASICs and something like a memory and CPU type of things. They did announce something in the market. And we should say this way, SMIC has been very closely work with these kinds of customers from the low end to the high end to the potential system players, and 12-inch to build out the technology platform to fulfill their requirements beforehand. And we are doing so on 3 parts. The first things first, we build up the technology platform, meet up their specifications and probably be, we are very strong in the mature nodes. But now, with Mong Song, [indiscernible] the R&D team speed up the technology development cycles, and we can quickly deliver the high-performance things to meet our requirements. The second thing that we build up capacity to meet these kind of additional requirements. The third for [indiscernible] side, we deliver the best cycle time and the competitiveness of PDKs, design manuals, IPs to make sure that we can meet both the power devices, BCDs, fingerprint, CMOS imager, PMIC, MCU, everything, this kind of demand growth -- 20% growth, they come from all fronts, later [indiscernible] requires both 8-inch and the technologies, IP, design manuals, PDKs, capacity, cycle time. Also the leading-edge technology, the new performance. So just now we mentioned that this year we increase the CapEx, a lot of -- majority of this kind of money will move into purchasing of R&D equipment to speed up the R&D also use this kind of equipment to do the Advanced Technology early production. And Mong Song, you can make the comments.

[Mong Song Liang, executive]

Okay. Yes, you said it all. Yes. Let me try to add some comments about you mentioned earlier about AI and cryptocurrency mining. Yes, indeed, we have confidence and we are preparing for the ASIC IP and to -- for this kind of new applications in first half of next year just to align with your early comments.

[Operator]

At this point, I would now like to hand the conference back to IR Director, Tim Kuo, for closing remarks.

[Tim Kuo, executive]

In closing, we would like to thank everyone who participated in today's call, and again, thanks, all of you, for your trust and support. Thank you.

[Operator]

Thank you. This is the end of SMIC's First Quarter Earnings Conference Call. We thank you for joining us today. You may all disconnect.