Michael Lin - Head of IR Jason Wang - President Chi-Tung Liu - CFO

Randy Abrams - Credit Suisse Bill Lu - UBS Michael Chou - Deutsche Bank Gokul Hariharan - JP Morgan Roland Shu - Citigroup Sebastian Hou - CLSA Rick Hsu - Daiwa Securities Julie Tsai - UBS

Welcome, everyone, to UMC's 2017 Third Quarter Earnings Conference Call. All lines have been placed on mute to prevent background noise. After the presentation, there will be a question-and-answer session, please follow the instructions given at that time. if you would like to ask a question. And for your information, this conference call is now being broadcasted live over the Internet. Webcast replay will be available within an hour after the conference is finished. Please visit our website, www.umc.com under the Investor Relations, Investors Events section. And now, I would like to introduce Mr. Michael Lin, Head of Investor Relations at UMC. Mr. Lin, you may begin.

Thank you and welcome to the UMC Conference Call for the Third Quarter of 2017. I'm joined by Mr. Jason Wang, the President of UMC and Mr. Chi-Tung Liu, the CFO of UMC. In a moment, we will hear our CFO present the third quarter financial results, followed by our President's message to address UMC's forecast and the fourth quarter 2017 guidance. Once our President and our CFO complete their remarks, there will be a Q&A session. UMC's quarterly financial reports are available at our website, www.umc.com under the Investors Financial session. During this conference, we will make forward-looking statements based on management's current expectations and beliefs. These forward-looking statements are subject to a number of risks and uncertainties that could cause actual results to differ materially, including the risks that may be beyond company's control. For these risks, please refer to UMC's filing with the SEC in the U.S. and the ROC securities authorities. Now, I would like to introduce UMC's CFO, Mr. Chi-Tung Liu, to discuss our third quarter 2017 results. Chi-Tung Liu: Thank you, Michael. I would like to go through the third quarter 2017 Investor Conference presentation material, which can be downloaded from our website. Starting on Page 3, the third quarter of 2017 consolidated revenue was NT$37.7 billion with gross margin at around 17.5%. The net income attributable to the stockholder of the parent was NT$ 3.47 billion. And the earnings per ordinary shares were NT$ 0.28 and utilization rates stay unchanged compared to the previous quarter at around 96% in the third quarter. On Page 4 is our statement of income on a quarter-over-quarter basis. Revenue basically stay flat around NT$37.7 billion with a gross margin around 17.5% or NT$ 6.6 billion. With contribution from non-operating income, our net income grew 66.7% quarter-over-quarter to NT$ 2.46 billion and EPS in the third quarter was NT$0.28 per share or NT$0.046 per ADS. On Page 5 is our year -- accumulative income statement. Year-over-year, we see a 2.8% revenue growth to NT$112.6 billion and gross profit was NT$20.7 billion and net income for the first 9 months of 2017 is NT$5.4 billion or 4.8 percentage point and EPS for the first 3 quarters NT$0.64 per share. So, page six is our balance sheet. And cash has been continued to increase to NT$69.9 billion at the end of the third quarter and total stockholder equity is NT$214.4 billion. On page seven is our operating segment report. Majority of our numbers and in fact over 99% of the revenue coming from our wafer fabrication department or foundry and new business segments now are very small portion of the consolidated numbers. So, on page eight, our third quarter blended ASP remained flat compared to the previous quarter. And for the revenue breakdown on page nine, euro is around 8% and the Japan is about 2%. For Asia and North America basically remained unchanged. On page 10, IDM account for 10% of our total revenue and the rest is fabless, is in-house. And on page 11, communication is still our largest segment, accounts for 47% of the total pie, with consumer remained around 31%. On the technology breakdown, page 12, we see 14-nanometer continue to be about 1% of our total revenue and 28-nanometer is about 15%. On page 13, we provide capacity details on a quarterly basis, and we still expect to see continued growth in capacity, mainly coming from Fab 12X in Xiamen. So, in the Wavetek Fab for both 8 and also 8S, we will also see some minor increase in capacity. And for CapEx in 2017, we have revised down to NT$1.7 billion in the previous quarter and now the number still stay the same, around NT$1.7 billion for 2017. And about this summary of the UMC results for third quarter of 2017, more details are available in the report, which has been posted on our website. I would now turn the call over to President UMC, Mr. Wang.

Okay, thank you, Chi-Tung. Good evening, everyone. Here I would like to update the third quarter operating results. In the third quarter of 2017, UMC's foundry revenue was NT$37.61 billion. In Q3 '17, we continued to sustain stable utilization rate across our 8-inch and 12-inch mature technologies, driven by a strong chip demand in consumer and computing peripheral segments. Our 8-inch facility remained nearly full while mature 12-inch fab operated above 90% capacity, lifting overall wafer shipment to 1.75 million 8-inch equivalents. Our 12-inch fab in Xiamen, Fab 12X, also began shipping 28-nanometer wafers performance reaching the same quality as our Tainan facility 12A. Looking into Q4 '17, we do expect the business environment to decline, due to the typical year-end seasonal adjustment. In addition, we foresee 28-nanometer High-K/Metal Gate demand to be soften. As we develop and redefine our new process technology according to market demand, we do expect to enhance UMC's market share by penetrating into emerging application including IoT, 5G wireless and industrial segments, which will spur new waves of growth opportunities. UMC's recent selection as a Dow Jones Sustainability Index global component for the tenth consecutive year and also highlighted our active involvement in environment protection and sustainable manufacturing practices, demonstrating our commitment towards setting higher corporate social responsibility standards. That was my third quarter results. Now let me go over the fourth quarter 2017 guidance. Our wafer shipments will show a decrease about 3% to 4%. ASP in U.S. dollars will decline by approximately 1%. Gross profit margin will be in the mid-teens percentage range. And capacity utilization rate will be in the high 80% range. For Foundy CapEx of 2017, it will be $1.7 billion. I will conclude my comments. Thank you all for your attention, and now we are ready for questions.

Thank you, Mr. Wang. And we would now begin our question-and-answer session. [Operator Instructions]. The first question is from Randy Abrams from Credit Suisse. Go ahead, please.

Yes, thanks. Good evening. Wanted to ask the first question on the CapEx. This year, year- to- date, you've spent about NT$1.1 billion. So, for fourth quarter, I'm curious if you still see the big spending hitting and where that's going to at the Xiamen fab. And if you could give an initial view on 2018 CapEx, if -- with 28-nanometer a bit lower if you can be more conservative if you have kind of a baseline view for next year?

We do expect CapEx would decline meaningfully in 2018. However, the actual number of the budget for the next year will be finished -- finalized probably towards the end of this year after the board's approval and we can share that after the -- in the next conference call. Chi-Tung Liu: For Q4 CapEx, this is more cash based, more related to payment schedule more than the capacity ramp. So, we do have some scheduled payment in Q4. So, most likely the number NT$1.7 billion is likely to still stay the same for 2017.

Okay. And the follow-up to that, your 12-inch utilization still was 80%, I guess a little lower in fourth quarter, could you give maybe an outlook for the 12-inch? Just how you see utilization in the next few quarters and also how you see the 28 businesses, it's 15% of revenue, if you could give maybe a percentage of revenue over next couple of quarters, maybe where you see it bottoming out on that side?

Let me just first address the 28-nanometer revenue contribution right now. For the Q3 '17, the contribution is about 16%. However, we do expect 28-nanometer High-K recovery will be challenging. So, and while we expect after the release of 28 HPC plus and 22 ULP platform in 2018 and, and we should see some recovery in the High-K area. For the Poly-SiON, the demand remain pretty robust and we know we have a high expectation of 28 poly-SiON area. So, that's in the High-K. The whole year, the utilization will be, all Q remain pretty healthy as a company expect 28 High-K. So, we think that in current year, the overall utilization growth will probably be in line with overall foundry industry.

Okay, maybe a view on current CapEx and OpEx. The CapEx, when you get that return in business on the 22 and also the 28 HPC plus, do you expect to or back to tracking operating cash flow. So once you refill the fab with those new applications, does the CapEx come back up or do you think sustainably become lower? And then I'm curious, the same on OpEx without the Investment in, say, 7-nanometer and a bit less aggressive on 14, if you now see after looking a few months, they give, you could do any either Cap up Ex could go even a bit lower?

Well, our CapEX strategies will align with our target of market segment and the growth target line. And I think at this point, we add a good level of the current market position. So our near-term hierarchy is to strengthen our financial flexibility and the cost structure. So, I don't see any, the changes in any time soon. However, it does change the market, in the situation, we'll update that. For the operating expense, right now what we do is we prioritize our R&D resource [indiscernible] as we reported last time and told to a specialty and a larger derivatives technologies. So, for the near term, we do expect operating expense to remain relatively flat.

Okay. And I guess, do you expect to sustain flatter, could you, if you don't need as much advance technology CapEx or do you think you have enough projects on specialty that have maybe kind of stayed in that flattish range?

Yes, I mean that's only on R&D side. I think we do prioritize those resources to specialty and the larger derivative line, the 22 HPC plus and 22 ULP, those will continue with the area we focus on. In addition to that, we also invest in the manufacturing productivity improvement as well as the quality assistance. So, those areas require continuous investment in both areas. So, at this point, in near-term, we expect to remain flat.

And next, we'll have Bill Lu from UBS for questions. Go ahead, please.

Yes, hi. Thank you very much. For the fourth quarter, can you also give us guidance on operating margin?

For the fourth quarter, the operating margin, I think at this point for the operating, probably we will try to ensure is operating profitable. For the profit margin wise, I think it will decline from the Q3, but I -- we are probably not going to comment on its actual percentage.

Okay. That's fair enough. I guess what I'm trying to get at is 4Q will be probably in the lowest single digits in terms of operating margins and typically 1Q is seasonally weak as well. Are there things that you can do to improve your breakeven levels in the next, say, several quarters?

It will be probably very challenging to improve in the shorter time, but we are doing everything we can to stay positive. And right now Q4 is challenging, but as far as going to Q1, we are too early to comment about this, but as far as the business involvement goes as we know, for seasonal reasons, Q1 will probably not likely to improve. And now only for seasonal reason, also because Q1 has a fewer working days and we do try to do some Q maintenance shutdown. So, that will also affect that.

Got it. A couple of questions on 28-nanometers. One is, I'm not sure if I heard you right, but it looks like High-K/Metal Gate is challenging, but I think you said that poly-SiOns remain actually pretty good, is poly up quarter-on-quarter in 4Q?

No, I mean the poly stay pretty flat but it is higher than what we expected originally. High-K is much weaken and so I would just say compared relatively, the poly-SiON remains healthier than High-K.

Got it. And then my second question on 28. Is -- you said that High-K demand will come back when you do HPC plus and 22-nanometers. I think last quarter you said that that will take 2, 3 quarters. Is the right timing for that to ramp back up still at the middle of next year? Chi-Tung Liu: Our expectation is on the 28 HPC plus and 22 ULP platform release. And the project is on track. The engineering project is on track for us to release the 28 HPC plus by mid-2018. And for the 22 ULP, that will probably follow -- be ready, we hope that would be ready 6 months after that.

Okay. So, when should we expect to see meaningful revenues from the HPC platform?

Our expectation is that we should see something in the second half of 2018 from the HPC plus and we will continue to see increase after the 22 ULP release.

Got it. And then one last question, which is a follow-up on Randys question. If I look at your R&D expenses, even if you don't cut hedge, the move from the leading edge to the specialty applications. I would assume that that would imply some cost [indiscernible] in terms of IP, in terms of tools, in terms of mass costs and things like that. Can you talk a little bit more about that?

Well, I think you're right. We have similar expectations, but we do have some intensive review in those areas. At this point, we actually spent a lot more on the net costs than we expected, because going forward, we probably have more customer engagement than the first way. We have one larger customer with higher volume. And now you have multiple customer engagement. So, it's probably lot more activities from engineering standpoint, maybe from no basis is less, but from a engagement point of view, probably more activities. So, that, associated with that, we actually had some increase in the R&D expense. So, but, we are cautious about that. So, we hope that we can see some meaningful reduction in the operating expense. But at this point, in the near term, we haven't really seen that yet. So, but that is an area we're going to continue focus on.

And the next question is coming from Michael Chou from Deutsche Bank. Go ahead, please.

Hi, good evening. Thanks for taking my question. My first question is what is the outlook by segment, can we say that your 8-inch remains still quite stable in Q4 versus Q3?

Okay, yes. I mean for the 8-inch, it is very stable and we continue to believe that will be in the full capacity situation. And so that's on the 8-inch. From the 12-inch, on the segment wise, we see stronger in the consumer area and this is the communication, as computer will get a bit weaker again in our portfolio.

So, you are referring 12-inch demand.

That's right.

Okay, fine.

So, the 8-inch that, for the 8-inch, you come from a various application very diverse and so far the demand remained pretty robust and healthy and so I think the 8-inch will remain full. And while we, in the 12-inch area, we do see the consumer has a much stronger outlook in the 12-inch while the communication and computer are really weaker, yes.

So, can we say your consumer segment could grow quarter-on-quarter while communication PC should decline quarter-on-quarter?

For the Q4, if you talk about outlook, actually reverse, what I'm referring to is that to Q3 results. And for the Q4, it's actually, we actually see a more seasonal effect in consumer and computer area.

Okay, thank you so much. Second question is a housekeeping question. What is the average currency for Q3 and what is your view for Q4? Chi-Tung Liu: Yes, 30 points is number for Q3. And Q4, just expect flat numbers.

Thank you so much. I'll back to the queue.

And the next question is from Gokul Hariharan from JP Morgan. Go ahead please.

Hi, thanks for taking my question. First of all, just wanted to explore a little bit on 28-nonometer. I think last conference call, I think you indicated, I think, almost 50% or more than 50% of revenue is poly-SiON. Could you comment about how that mix looks like in Q4? And also, could you talk about how broad-based this demand is and sustainability of poly-SiON demand into first half of next year? Just wanted to understand that because you mentioned that the HPC plus revenues will start coming through only from second half of '18. That's my first question and I had a follow-up as well.

Alright. So, the first, for the poly-SiON, split between High-K and poly-SiON. The poly-SiON stays about 50%, say, in the fourth quarter and we do see that remains firm. So, they will probably stay in that range.

Even into first half, you think that that should be pretty sustainable.

I [indiscernible] to continue with the projection in the first half, but because the 28 High-K is softening, we do see that actually will be more than 50% in the first half with the poly-SiON.

Okay. Could you talk a little bit about the longevity of the poly-SiON projects, because for some of the other foundries, there is an accelerated conversion from poly-SiON to High-K/Metal Gate and derivatives of High-K/Metal Gate? So, what do you see from your customer base, is it like multiple projects on poly-SiON with longer duration that you're seeing or...?

Well, yes, we do see that and we continue to see engagement in the poly-SiON area, although we have a lot of engagement in the High-K as well, but because of timing reasons and so the High-K will probably not coming in until later part of 2018, but for the [parties], so far we haven't seen any weakness yet.

Okay. And one question for Chitung. On the depreciation trend, could you talk about what do you expect on depreciation in Q4 and maybe you have some indication going into next year as well, is depreciation still going to be rising into next year? Chi-Tung Liu: For quarterly basis, it's rather flatish, Q2, Q3, Q4. Q3 is probably up a little bit, but all 3 quarters are similar. And in terms of year-over-year basis, 2017 and 2018 based upon current projection should be quite flat. And I guess we've talked about before, we do expect to see some kind of decline in depreciation or at least pick-up in terms of depreciation on an annual basis starting from 2019.

And the next one is from Roland Shu from Citigroup. Please ask your question.

I think first of all, operating expense, last time you said going forward for the operating expense, it will be back to normal and this time earlier you said, now you expect operating expense to be flat in the near term. So, can we have a more clarification for this operating expense. And when you talk about back to normal, what do you mean the normal label for this operative expense? Chi-Tung Liu: Again, I mean basically OpEx throughout will stay at a current level as a percentage of revenue or absolute dollar terms. In the near term, we don't expect to see too much downside on the operating expenses, even though we see potentially for the longer term, we will be able to drive down the operating expenses and also hopefully couple with the revenue growth, percentage of revenue wise for OpEx can gradually start to decline in the longer term. But that's not going to happen in the near term.

So, Chitung, you mean the quarterly operating expense number, right? Chi-Tung Liu: Yes, quarterly, yes.

And in 3Q, you had a quite significant gain up from down up. Can you add more color on this on up gain in 3Q? Chi-Tung Liu: Third quarter is normally in the season where we see dividends from investing companies, so that is always the case. So, some NT$500 million dividend we'll receive in Q3. Also, we benefit from the reversal of the strengthening renminbi given our U.S. fab in Xiamen fabs. So, that's another NT$400 million, NT$500 million contribution from the reversal of forex gains, which would book us loss in 2016. So that's the 2 major items.

And for your maintenance CapEx, how much it is for every year for the maintenance CapEx primarily? Chi-Tung Liu: It will be probably in the range of NT$200 plus million.

NT$200 plus million. So, going forward, the capacity expansion in China, what is maintenance CapEx increase going forward? Chi-Tung Liu: No, if you talk about next 1, 2, 3 years, I don't expect the number to vary differently. So, it will be in this ballpark figures in the recent years.

And last question is on your, for your 28-nanometer, revenue was about 15% of the total in 3Q, how much came from Taiwan and how much of this 28-nanometer revenue came from China in 3Q?

28-nanometer in terms of capacity in our Xiamen fab, currently only around 5K, in Taiwan we have more than 35K. So, capacity wise, it's 1-7. And revenue wise, it just started in Xiamen. So, it will be even less than the capacity ratio.

Next, we'll have Sebastian Hou from CLSA for questions. Go ahead, please.

Hi. Good afternoon everyone. So my first question is, can you remind me of your capacity split on 28-nanometer in terms of poly-SiON and High-K/Metal Gate?

We have by end of this year and we have approximately 39.3K of 28-nanometer and between the two, the poly-SiON has s about 13K while the High-K has [indiscernible].

But there is some kind of conversion.

Right. Well, in between, there is -- we have a conversion ratio, we have about 7K to 8K that we can convert between the two. So, that's -- when we have poly-SiON stronger, we're actually using the High-K capacity for the quarter, the [poly-SiON].

So, does that means that when you -- so, it could be the maximum about 20K of poly-SiON?

Yes, 20K to 21K, yes, in the range of it.

If you convert everything, you can convert?

Yes, convert.

Okay, thank you. Can you update us on the -- if you look at your 40-nanometer, which has been pretty strong and continue to be very strong, can you share with us what kind of applications to drive that and how sustainable is that?

Well, you're asking the 14 or the 40-nanometer?

4-0. Chi-Tung Liu: 4-0, 40-nanometer?

Yes.

I mean, this -- there is actually a variable different application into 40, and from the logic to the high voltage and battery fresh. So, there are some specialty as well as the larger device and for the platform of the mobile devices and also the MTU despite private areas.

And we have noticed that you are -- you have expanded and you are also expanding your age capacity into the second half of this year. So, I think after several quarters, with the flattish expansion, so what makes the changes and also how much room -- how much idle clean room space you have for further expansion into 2018?

For age?

Yes.

Very limited clean room space for expansion due to the demand, robustness. So, we are trying to continue to expand a little bit of the age for the older space that we have.

So is it fair to assume that even into 2018, given the limited clean room space you have, you probably will not -- even you add, will not be much?

Right. It won't be much, but they may have some of the product mix adjustment due to different products.

Okay, thank you. In terms of the earlier, Chitung, your guidance about depreciation to be flat in 2018, does that already factor in the meaning from CapEx cut that the President guided earlier? Chi-Tung Liu: Yes, that's the current projection.

Last question from me is that, do you have any tape-offs already on 28 HPC plus right now and can you share how many you have?

No, I mean we don't have any HPC tape-off yet, no.

Okay. So, when you say the technology is ready by mid of next year, does that mean that the same timing for tape-off or as your tape-off will likely start earlier than that or pretty much the same time, mid of next year?

For those customer require the IP, we probably won't be able to take tape-out until the 2018. For those customer does not need IP and we can try to tape off early than that.

[Operator Instructions] And next we'll have Rick Hsu from Daiwa Securities for questions. Please ask your question.

The first question is on your 28, sorry, 22-nanometer ULP. Can you clarify, is this also based on High-K/Metal Gate poly-SiON?

This is not High-K process.

And do you have any tape-off right now?

No. The 22 ULP will not get ready until mid of 2018.

Second question. Can you comment on your customers' inventory status?

For seasonal reason we do anticipate some of the inventory correction across the board starting in Q4. And, but we're not commenting specific customers. But we do starting some of the inventory correction activities.

Last question is, do you have any revenue contribution from cryptocurrency mining?

The answer is yes, and we have some of the demand coming from that segment.

And would you mind sharing with us is based on what technology note and how do you see the trend going forward?

Well, because it's for specific customer and [indiscernible] space is always, we will not comment about that.

And the next question is coming from Randy Abrams from Credit Suisse.

Yes, I had 2 follow-up questions. The first one on the specialty technology that you're reallocating the resources. Could you consider if there is a few certain product or areas that you're targeting like that you think you can grow the business like you've had the specialty in the past, but maybe now with the new resources which products or area if you see the most promised?

I mean, those areas that we currently focus is a bit sensitive and so, and they may actually reflect to certain customer engagement, so I kind of reluctant to report that, but we are focused on some of the specialty areas that's going to market demanding, and we feel pretty good traction at this point.

Maybe if I could ask, I guess, when you think, I guess, it's kind of a new effort to reallocate when -- like if you might see an acceleration in some of these areas. And if there certain nodes that you're seeing the applications, I guess, 8 inch to slow, so is it like some of the mature, like could you expect it's more [40s], like which nodes you expect and when do you think you'll start to see the contribution?

The contribution is already started in some area and especially in 8-inch area and we spend many efforts in the 12-inch mature space. Now hopefully that we can continue to migration from 8 inch to 12 inch, so it covers between the 8 inch equivalent and the mature 12 inch.

I had 2 P&L clarifications for Chitung, one is there is a line item net other income that flows into OpEx and it has been like plus to minus NT$ 400 million to NT$ 500 million and this past quarter was a plus NT$ 441 million. If there's a way to think of what that other income is and if you expect that to sustain and the other line item with the non-controlling interest, how you see that, that's about NT$ 1 billion now, how you see it trending and Xiamen ramps up? Chi-Tung Liu: Non-controlling for just the consolidated accounting principle at a pace the currently the loss of Xiamen fab to other shareholders. And that numbers are percentage at least will decrease when UMC continue to increase our shareholding in the Xiamen fab. As for the other income in many of the subsidiaries we receive from our investment in China.

Do you think it will be similar, roughly similar in the next few quarters or... ? Chi-Tung Liu: That's a bit difficult to predict, but I assume the number will be above the same next quarter.

Okay, great. Thanks a lot Chitung.

And the next one is from Julie Tsai from UBS.

Hi, I have a follow-up question from my analyst, Bill Lu, regarding to the operating expenses. I think, Jason, you didn't mention that in the near term your mission for UMC is actually to strengthen the financial capability and cost structure. So is it safe to assume that the operating expense will probably maintain around NT$ 5 billion per quarter just like last quarter or should we be using a percentage which varies from, I would say, quarter 1 to the most recent quarter, it could be 16% of sales or 13% of sales?

I mentioned that in the previous questions that we are trying hard to exempt the structure of our operating expenses. In the near term, we don't expect to see any significant or meaningful reduction either in absolute terms or as a percentage of revenue. But for the longer term, of course, we do expect to see more rationalized in the operating expenses coupled with our increasing revenue base. So at least for the percentage of revenue, hopefully for the longer term, we hope that as a percentage of revenue will gradually decline.

Because I think the sell side community uses recent quarters from operating expenses of NT$ 5 billion, then I think the next few quarters, operating income would be okay or pretty much sustainable at a current level, but if they use your percentage of the sales, it could actually maybe come into very near breakeven, would that be the case? Chi-Tung Liu: I cannot comment on the first quarter of next year but we already mentioned that Q1 normally is seasonal slower quarter coupled with our maintenance schedule and fewer working days and then we will talk about our operating outlook for Q1 in next quarter.

And then maybe just follow-up, I think, Jason also mentioned that 2018 CapEx was the meaningful decline. Is it okay to quantify what's meaningful, is it 20%, 50%...

There is a reason what gives meaningful, so no, we cannot.

Not yet, right, okay, we will leave that to maybe in next quarter.

Sure.

And the next question is coming from Sebastian Hou from CLSA. Go ahead please.

I have 2 follow-ups, very quick follow-ups. The first one is on your 28 nanometer capacity plan and your Xiamen fab expansion plan, do you plan to expand further capacity on 28 and also in Xiamen next year?

Yes, and by end of this year. we will have about 11.5K total capacity in Xiamen and we do expect that number will reach about 16K by mid-2018. And most of all the addition is 28-nanometers.

It's 28 poly or 28 High-K?

Poly.

Poly? Okay. So, how about your Tainan fab 28 in Taiwan?

At this point, there is no expansion plan planned up for the 12A yet.

And on the, my second follow-up is on the DRAM that the JV you have with change on government, if I read about it correctly, so can you give us some update on that, when do you expect the wafer start to begin? Chi-Tung Liu: Well, first of all, it's not the JV, we only are technology provider, so you will see us probably investment in this DRAM initiative. And Jason, can you comment on [indiscernible]?

Again, just like Chitung said, personally you're right about the location, it's Jinghong in China and close to Xiamen. The R&D activity is on track at this point and we do expect to see some later output, the engineering wafers by the beginning next year and in our development facility in Taiwan and the facility in Jinghong is under construction and so far the project is still on track. Just like Chitung said, we actually don't own anything in the project. We're more of a technology partner and to support that project.

Okay, thank you. So, engineering wafer is starting in 2018 and the real output will be...?

Well, Jinghong's real wafer output is really up to their schedule. I understand the production facility is under the construction and the project is still on track. But I really can't comment when they're going to have a production out, because it's subject to when they're going to have the tool move in. But from a development point of view, the project is on schedule.

Okay. Rhank you very much.

We thank you for all your questions. That concludes today's Q&A session. And now I'm turning things over to UMC Head of IR for closing remarks.

Okay, thank you for joining us today. We appreciate your questions. As always, if you have any additional follow-up questions, please feel free to contact UMC at ir@umc.com, and have a good day. Thank you.

Thank you, Mr. Lin. And ladies and gentlemen that concludes our conference for the third quarter 2017. Thank you for your participation in UMC's conference. There will be a webcast replay within an hour. Please visit www.umc.com under the Investors Events section. You may now disconnect. Goodbye.