Ladies and gentlemen, thank you for standing by, and welcome to Hua Hong Semiconductor's third quarter 2020 earnings conference call. Today's call is hosted by Mr. Danny Wang, Executive Vice President and Chief Financial Officer. Be advised that your dial-ins are in a listen-only mode. However, at the conclusion of the management presentation, there will be a question and answer session, at which time you will receive instructions on how to participate. The earnings press release and third quarter 2020 summary slides are available to download at our company's website, www.huahonggrace.com. Without further ado, I would like to introduce you to Mr. Danny Wang, Executive Vice President and Chief Financial Officer. Thank you.
Good afternoon, everyone. Thank you all for joining our third quarter 2020 earnings conference. Unfortunately, Mr. Tang, our President and Executive Director, has a meeting at this time, which require his presence. He has asked me to express his disappointment at missing this call and to cover our third quarter performance on his behalf. Today, I will begin with President's comments on our third quarter performance. After that, I will discuss our financial results and provide guidance for the next quarter. This will be followed by our question and answer session. We are very pleased with Hua Hong Semiconductor's performance in the third quarter of 2020. The company's sales revenue hit a new high, achieving double-digit quarter-on-quarter growth for two consecutive quarters. Although COVID-19 has not been fully controlled worldwide, the prosperity of the semiconductor industry has gradually improved as domestic consumption has rebounded significantly, and a robust end-user demand has continued.
Driven by a strong demand for MCUs, IGBTs, logic chips, and other products, the company achieved historical record revenue of $253 million. 8-in capacity utilization continued to be strong, exceeding 100% for the three 8-in plants in the third quarter. 12-in capacity utilization also increased significantly. Benefiting from high capacity utilization, gross profit margin exceed our guidance, and profitability has been achieved for 39 consecutive quarters. Introduction of new Hua Hong Wuxi 12-in products is progressing steadily in accordance with our process development plan. Yield improvement production capacity ramping are proceeding much faster than our original plan. Currently, we are in mass production and are shipping products from embedded flash memory, logic, and RF, and low voltage power device platforms. In the fourth quarter, more new products will be delivered for customer verification, such as IGBT and superjunctions.
The focus of our next phase of work will be to accelerate capacity expansion in Wuxi to alleviate the current limitation of 8-in capacity and to expedite development and improvement of our process platforms to provide more comprehensive and better product solutions for our customers. In the face of the second wave of COVID-19, we must firmly maintain the results of our pandemic prevention and control measures, keep close communication with domestic and foreign vendors and customers, actively expand our markets, and deepen our cooperation with all stakeholders. I would like to thank all shareholders, customers, and vendors for their support, as well as all employees for their relentless efforts. Although the road ahead will always be full of challenges, we are confident we will be victorious with our down-to-earth efforts. We believe we will continue to deliver more outstanding results in the future.
Next, let me begin with the summary of our financial performance for the third quarter, followed by an outlook on revenue and margin for the fourth quarter 2020. Then we will move on to the question and answer session. First, let me summarize our financial performance of the third quarter. Revenue hit an all-time high of $253 million, an increase of 5.9% over the prior year, and 12.3% over the prior quarter, mainly driven by increased wafer shipments. Cost of sales was $191.7 million, 16.2% above Q3 2019, and 14.9% over Q2 2020, primarily due to increased wafer shipments and depreciation expenses. Gross margin was 24.2%, 6.8 percentage points below Q3 2019, and 1.8 percentage points below Q2 2020, mainly due to increased depreciation expenses and other fixed costs from the new plant in Wuxi.
Operating expenses were $74.2 million, 84.7% above Q3 2019, largely due to increased development costs and reclassified fixed costs from the new plant. And 18.7% over Q2 2020, primarily due to increased development costs and labor expenses. Our income net was $24.1 million, 2.3% of associate, partially offset by decreased fair value gains on financial assets at fair value through profit or loss and interest income. And 100.5% over Q2 2020, mainly due to increased government subsidies and a foreign exchange gain versus a foreign exchange loss in the previous period. Income tax expense was $10 million, $22.3 million lower than Q3 2019, primarily due to decreased taxable profit. Profit for the period was $1.1 million versus $44.4 million in Q3 2019 and $1.3 million in Q2 2020.
Net profit attributable to shareholders of the parent company was $17.7 million, compared to $45.2 million in Q3 2019 and $17.8 million in Q2 2020. Basic earnings per share was $0.014 versus $0.035 in Q3 2019 and flat to Q2 2020. Annualized ROE was 3.2%. Now I discuss the operating results of both the Hua Hong 8-in wafer fab and Hua Hong Wuxi 12-in wafer fab. First, let's have a look at the Hua Hong 8-in wafer fabs. Revenue was $236.4 million, 1.1% below Q3 2019 and 9.5% over Q2 2020, mainly due to increased wafer shipments. Gross margin was 27.2%, 3.8 percentage points below Q3 2019, mainly due to increased depreciation expenses, partially offset by improved capacity utilization and a 0.5 percentage point below Q2 2020.
Operating expenses were $30.8 million, 14.5% over Q3 2019 and 16.9% above Q2 2020, primarily due to increased labor and depreciation expenses. Profit before tax was $45 million, 23.7% below Q3 2019 and 7.7% over Q2 2020. Now let's have a look at the performance of Hua Hong Wuxi wafer fab. Revenue was $16.6 million, 75% over the prior quarter. Operating expenses were $43.4 million compared to $13.3 million in Q3 2019 and $36.2 million in Q2 2020, mainly due to increased development costs. Profit before tax was -$33.9 million. EBITDA was -$11.5 million, $7 million more favorable than Q2 2020. Now I would like to provide more details on revenue from Q3 2020. From geographical perspective, revenue from China was $165.3 million, contributing 65.4% of the total revenue, and an increase of 11.2% over Q3 2019, mainly driven by increased demand for logic products.
Revenue from the United States was $34.1 million, a decrease of 3.4% compared to Q3 2019, mainly due to decreased demand for superjunction products, partially offset by increased demand for MCU products. Revenue from Asia was $30.4 million, an increase of 2.7% compared to Q3 2019, chiefly driven by increased demand for logic products. Revenue from Europe was $16.3 million, a decrease of 4.9% compared to Q3 2019, mainly due to decreased demand for smart card ICs. Revenue from Japan was $6.7 million, a decrease of 17.2% compared to Q3 2019, mainly due to decreased demand for logic products, partially offset by increased demand for MCU products. With respect to technology platforms, revenue from embedded non-volatile memory was $87.3 million, a decrease of 0.9% compared to Q3 2019, primarily due to decreased demand for smart card ICs largely offset by increased demand for MCU products.
Revenue from discrete was $96.2 million, an increase of 6.6% over Q3 2019, mainly driven by increased demand for general MOSFET and IGBT products, partially offset by decreased demand for superjunction products. Revenue from analog and power management IC was $33.5 million, a decrease of 8.2% compared to Q3 2019, mainly due to decreased demand for analog products. Revenue from logic and RF was $32.7 million, an increase of 48.5% compared to Q3 2019, largely driven by increased demand for logic products, partially offset by decreased demand for RF products. Revenue from standalone non-volatile memory was $2.8 million, an increase of 59.8% compared to Q3 2019, primarily driven by increased demand for flash products. Now let's take a look at the cash flow statement.
Net cash flows generated from operating activities were $89.4 million in Q3 2020, compared to $70.7 million in Q3 2019, largely due to receipts of VAT tax return, partially offset by increased payments for labor. Capital expenditures were $403.6 million in Q3 2020, including $351.2 million for the Hua Hong Wuxi fab and $52.4 million for Hua Hong 8-in fabs. Other cash flow generated from investing activities were $88.2 million in Q3 2020, including one payout of $86.9 million from investment in financial assets at a fair value through profit or loss, and two, $1.3 million of interest income. Net cash flows generated from financing activities were $222.1 million in Q3 2020, including, one, $219.5 million of proceeds from bank borrowings. Two, $3.7 million proceeds from share option exercise, partially offset by $1 million of lease payments and $0.1 million of interest expenses for bank borrowings.
Now let's move to the balance sheet. Cash and cash equivalents was $716.5 million on September 30, 2020, compared to $699 million on June 30, 2020. Financial assets at a fair value through profit or loss decreased from $84.8 million on June 30, 2020 to zero on September 30, 2020, due to payout from investment in financial products. Inventories increased from $176.9 million on June 30, 2020 to $205.2 million on September 30, 2020, primarily due to increased wafer demand from customers. Property, plants, and equipment was $2,272,500,000 on September 30, 2020, compared to $1,790,900,000 on June 30, 2020. Total assets increased from $3,575,900,000 on June 30, 2020 to $4,022,700,000 on September 30, 2020. Our total bank borrowings were $243.7 million on September 30, 2020.
Total liabilities increased to $848 million on September 30, 2020 from $526.1 million on June 30, 2020, primarily due to increased bank borrowings and payables for capital expenditures. Debt ratio increased to 21.1% on September 30, 2020, from 14.7% on June 30, 2020. Finally, let me give you a top-level outlook for the fourth quarter 2020. We expect revenue to be approximately, excuse me, approximately $269 million, and our gross margin to be between 21% and 23%. This concludes my financial remarks. Now we would like to start the question and answer session. Operator, please help. Thank you.
Thank you. Ladies and gentlemen, we will now begin the question-and-answer session. If you wish to ask a question, please press star one on your telephone and wait for your name to be announced. If you wish to cancel your request, please press the pound or hash key. Again, it is star one if you wish to ask a question. Your first question comes from the line of Leping Huang from CICC. Please ask your question.
Okay. Thank you for taking my question. I have two questions. The first question is for your Wuxi fab. Can you share some color on the customer or the ramp-up progress going forward? Especially for us, we see the third quarter, your revenue already grew very strongly. You see that. What are the drivers and what are they for the future additional application sensor? The second question is about your 8-in. We see that the 8-in are in a shortage status. I think the market is expecting that the ASP of the 8-in fab can go up. What is your status now? Do you think you can also expect the ASP for the inch fab to go up? Thank you.
Thank you, Leping. Very good questions. I've been getting them lately a lot. First of all, on the Wuxi fab. Okay. Let me just go through with you the technology platforms that we're currently working on. As I said many times in the past, we have virtually six different technology platforms that are going through R&D work or customer verification process. Okay. Or just going through that process. Okay. We have the logic and RF, okay? It's the 55 nm RFCMOS. Eventually, we get to RFSOI. Okay? The RFCMOS, it's already in production. Okay. We're planning for more new products. RFSOI, its production has not started, but we expect it will start up probably sometime next year. Most likely, mid of next year. Okay. CIS, 90 nm CIS. Okay. At the frontside illumination, we're talking about, it's the 2-MP. Okay?
This is the main driver now for the Wuxi fab. We are talking about 10,000 wafers loading at this point on a monthly basis. A great demand for cell phone and other consumer applications. Our plan, it would continue to ramp up for this product in Q4. We expect that there is going to be a huge revenue jump in Q4 just from this product. The other thing is the 65 nm backside illumination. This is something we are going to be doing sometime next year. Probably we will start Q1 2021 on the 8-MP. Eventually, the 16 MP and 64 MP. The other thing is embedded non-volatile memory. We have already started production for the SIM cards and some bank cards. Early stage, but we are talking about 1,000 loadings, mostly for the SIM cards.
Eventually, I think this thing will get to about 3,000. We have said in the past we are going to migrate most of the smart card ICs from the 8-in to 12-in. So this is something we definitely will do. It is going to help us, help our customers. It is not the best ASP product, but by moving from the 8-in to the 12-in fab would definitely help. It is going to be a win-win situation for us as well as for customers, because it will make the cost more efficient. Should make the ASP more stable. So long term, 12-in definitely is the thing. Eventually, we are going to do that in a 55 nm platform.
But this will be something like, I would say, the 55 nm, where probably we are going to start some new tape out in Q1, and probably the production will start on second half next year. Eventually the MCU, but it is still early stage. We are going to start work on a few products now, and then I think the production will really start later because we have a lot of space in our 8-in fab after we migrate the smart card ICs from the 8-in to the 12-in. Therefore, we can do a lot of MCUs. You all know that MCU is a very high margin product, very stable. The growth has been just tremendous this year. We are talking about MCU itself is probably going to be over 20% year-on-year. Many applications. The other thing is NOR flash.
90 nm NOR flash. We are going to be probably start production next year. Very solid customer we have in China. You have 16 Mb, and eventually 32 Mb. So that is something we are going to be doing. Then, of course, we eventually will move on to 55 nm. So that is going to be the plan. Then power management IC, 90 nm BCD. We have a very good international customer who have signed a strategic agreement with us. We are starting the risk production now, have already started. We will start to ramp up really starting, I think, Q3 and Q3. This is for eventually fast chargers, that sort of things. But this is going to be a very major product for us in the future. Finally, the power discrete product.
As a foundry, we have the biggest capacity for power discrete, okay? Globally, nobody has a bigger foundry discrete capacity than Hua Hong Semiconductor. Many good customers in China. We have already started production. Currently, I think we are doing a lot of DMOS, split gate trench. The other thing is Deep-Trench superjunction. Eventually we are going to be doing the IGBT. I think we are going to be doing mostly the IGBT, the high voltage stuff in this 12-in, the fab, okay? We have about 3,000 a month now in production, already 3,000 power discrete in production. Certainly, it will be more and more, okay? Just overall, the strategy is, I think, our ASP. People were talking to me earlier, had some questions, say, "Oh, your 12-in fab ASP has gone down." That is absolutely not true. The ASP for 12-in has not gone down.
In fact, very stable, even has a slight increase quarter-over-quarter compared to past quarter, okay? At this point, it is very important. We are moving very quickly with this fab. We have 20,000 wafers capacity, okay? By end of this year or early next year, it will get to about 40,000 wafer capacity, okay? The goal is to get to production. We are loading at this point, even though the utilization rate for the Q3 was 50+% . The reality is just it is currently already loading about 15,000 wafers out of the 20,000. So we are talking about 70% utilization rate. It has already gone up, okay? The goal is to get to about 20,000 wafers in production or shipment by end of this year. At the same time, we are going to expand the capacity to about 40,000, okay?
Hopefully get to 40,000 by end of this year and early next year. So we will have 40,000 wafers virtually to load, okay? The goal is to get the 40,000 wafers completely loaded by end of next year. This is the capacity loading plan, expansion plan, okay? We will quickly load this 12-in fab. At this point, as I said, we have six technology platforms. Three of them are already in production. The key is to make sure to get this capacity loaded very fast and almost fully. So make sure the utilization rate is very high. Therefore, you can reduce the fixed cost, okay? That is the most important thing. Initially, it is very critical to make sure this fab is loaded.
From that point on, with the additional technology platform, we are going to improve the customer mix as well as the product mix, okay? You are going to see, from this point on, I think the ASP will gradually improve. A portion of that will be loaded power discrete. Power discrete, because it has less manufacture layers, so the ASP will be just in general is lower, okay? It all depends on mix of ICs versus power discrete. That is the strategy. Overall, the ASP should only improve. You can see, look at the data, it is about $1,000 at this point overall, the average ASP, okay? I think from this point on, it can only go upward. That is the plan for the 8 in, 12 in. As far as the 8 in it is 8-in is running very, very full.
At this point, our 38-in fab is running anywhere between 100% to about 110% for Q3. Extremely high. I think we are able to improve the product mix a bit. The ASP actually did go up in Q3 slightly. We did not really have a policy that going out lift our ASP. It is not our strategy. These things will be handled seamlessly if the demand continues to be strong. I think the ASP will get improved anyhow because of customers' requirement and also product mix improvement. This is something we would do, but I surely expect that our ASP will go up. I think we should be able to improve in Q4 somehow as well. Let us see how that goes. But overall, we are doing it. The loading is very, very high.
Definitely the second half is much stronger than the first half that we experienced, especially we came out the Q1, which was really a low note for us. But where we pass that point, we are really in a great shape at this point. With the strong demand currently, I think things can only get better. Thank you, Leping.
Thank you. Your next question comes from the line of Randy Abrams from Credit Suisse. Please ask your question.
Okay. Yeah, thank you. Daniel, a couple follow-up questions to the prior. Just for the 8-in running over 100%, do you have any plans to de-bottleneck and add capacity at this stage, or is this solely the Wuxi fab?
For the 8-in fabs, you are talking about, Randy?
Yeah. If you plan to add any, I think you have room for a bit more, but if you plan to do that or just keep it where it is.
Yeah, absolutely. We are spending some money throughout the year, we have been investing. We continue to invest in the 8-in space. I think the investment is about, as far as the CapEx, is around about slightly over $100 million, which is very normal for these three 8-in fabs. In fact, it is probably still under sort of normal level. You can see actually our depreciation expense also went up a little bit because some might say why your growth margin was affected because we have been investing. We continue to invest in the 8-in space. So, in fab one, fab two, as well as fab three. I think with these investments, the three 8-in fab will continue to become better and better. It is very important as we are ramping up in Wuxi, we need to make sure we have a solid three 8-in fabs that are doing well.
The timing is in our favor because market is very, very strong. As I said earlier, we will continue to do that. We also allocate some money budget for next year as well. We just had our Board meeting. They are very, very supportive of us doing that. I think the three 8-in fabs, I think we want to eventually get back to this 30+% gross margin that is important for us.
Okay. To follow up on that, the net capacity, would that change or is that fixed? You're doing technology upgrades, but would there be any expansion in the actual wafer capacity you can do through these upgrades?
Well, the upgrades we're doing is virtually doesn't really change the adding the capacity for any of these three fabs. You know what? We do have some space in our fab three. We can add about, depending on, we have to think about this. At one point we decide to maybe we should do some basically expand capacity for the power discrete. But we have to make a decision. We still have to think about that. We can add about anywhere 20,000 wafers capacity for power, which for regular IC probably is going to be around under 10,000. So this is the space we have. If the timing, if we do have that need, if we have strong demand, then we probably would do that. Certainly we have enough budget to cover that if we want to do that for next year.
Okay, great. For the CapEx, could you maybe just give an update where this year would come in for CapEx, then initial view on 2021 as you finish phase one and start to consider the next phase, just if there's a view where CapEx could be.
Yeah. You're talking about specifically for 8-in or for Wuxi. Let me just give you an overview, okay?
Yes, if you could.
This year, on cash flow basis, the 8-in business, we're basically less than two months away from the year-end. I expect these numbers, it's getting very close to the actual number. I think, I'm talking about close to $160 million capital expenditure we spent for the 8-in fabs, okay, three 8-in fabs. I think probably it's going to be around $70 million expansion, additional capacity, just the bottleneck tools, that sort of upgrades. Then we're also about $46 million maybe on maintenance, just fab maintenance, okay. That's for the 8-in. Now, for the Wuxi fab, it's a $2.5 billion investment this year, I mean, for the project. But this year, I think on cash flow basis, I think it's going to be about $1.25 billion, okay? That's probably, it's going to be the spending.
Of that, there's about $100 million where we're building the dorm for the technicians, the employees. There's that part, but the rest of it's about close to $1.1 billion + that goes to the, basically mostly for the tools, okay. Now, for the $2.5 billion project, there's still about another $300 million remaining. I think the tools have already been ordered. I think it's just the timing is probably going to be next year for those payments, okay? I think next year, the remaining balance will be another $300 million, again, mostly for the tool, okay? Because of demand is very strong, okay, we're thinking about next step, okay? We're thinking about next phase of the expansion. We have built a huge fab. It's a mega fab, as I mentioned in the past. Right now, it is 40,000, for the first initial phase.
But the fab itself really can handle about 80,000 wafers. We're already start to think about the next phase, okay? We probably have to start the basically, we probably can do this in two stages. We can add 20,000 and then another 20,000. That would require some additional investment, definitely. We're looking into that.
Okay. It sounds like next year, factoring $300 million, 8-in probably continues near similar. Next year, it sounds like some work in process, but could still be a decent amount of spend if you move ahead with the second phase.
Yeah, definitely. Well, it's another 40,000 wafer capacity. I think it'd be going to be smaller than the initial phase, okay, the first phase. Now think about this, okay? If we have 80,000 wafer capacity, we did the calculation, we did the model, we can basically, this fab will generate about $1 billion revenue, just one fab, okay? It's going to be very close to what we have from the three 8-in fabs, okay? Currently, the 3-in fabs is getting very close to $1 billion, okay? Hopefully, we can get to that number next year, okay, and that'll be fantastic because, at the same time, the 12-in fab will next year will also have a target for that, okay, in terms of revenue. I think it'll be fantastic. I would expect we definitely will move forward, continue to expand.
Hopefully in the next two to three years, this fab will generate another additional $1 billion for us.
Okay, great. The final question I have is just how are you thinking then about the break even on the first phase, and the profitability as you ramp up the second phase? If you could also provide the depreciation.
Yeah, definitely. Well, I always said, please focus on the growth, on the revenue this year, next year. Just pay attention to growth of the revenue. That is very key for us. The EBITDA will come with it. Look at how well we did this past quarter. Virtually, I think giving another quarter or two, I think we should get to EBITDA positive very easily. Okay? So that gap was reduced by about $7 million, $8 million this past quarter. Very significant. Okay? The EBITDA was about -$22 million. Okay? For Q2. Now it is reduced by $7 million, more than $7 million, in fact, it is now - $13 million, so it is reduced by $9 million. Okay? So, pretty soon, I think in probably another quarter or two, that we can get to EBITDA positive. And then hopefully, we are ramping from 20,000 to 40,000 by end of next year.
That will get us to sometime, hopefully, next year, second half, we can achieve that. It is not easy, okay? Because we are investing a lot of money in R&D, okay? And also the depreciation expense is going to be very high, okay? So, if we can get to that, we are probably going to set a record, okay? But, I am hoping that by end of next year, we can somehow get very close to that, okay? That is going to be important. So on your next question, what was your next question, Randy?
That was just on the depreciation, if you have a rough-
Oh, yes.
number for next year.
Yes, definitely, on depreciation expenses. The depreciation amortization for the 8-in fab is expected to be around about $130 million for the three 8-in fabs. For the 12-in fab, it is about $81 million. Overall, I would say it is about $220 million-ish, somewhere $210 million-$ 220 million. This is how we're going to end the year for depreciation expenses for this year.
Okay. Is it too early for next year?
No. We have to have these numbers. It may not be accurate. Things may change. Could be more, could be less, but, I certainly have a plan for that. You want to know about maybe just a rough number for next year?
Yeah, if you have that. Thank you.
Okay. One second. Give me a few seconds here. Oh, okay. Right here. Let me take a look at this. I would expect for the 8-in business, I think for next year, it would probably be around pretty close to what they have this year. Okay? It is about $130 million - $140 million. Okay? But we like to keep in that range, okay, $130 million - $140 million. For the 12-in fab, I think that is going to go up. Okay? This year, I said earlier, it is going to be around $80 million. I think next year, we will probably have to double that number to about $160 million - $180 million. Okay? That would be the number. Overall,
Thank you, Daniel.
We are looking at overall about $300 million - $310 million.
Okay. Great. Okay. Thanks, Daniel, for the update.
Thank you. Your next question comes from the line of Szeho Ng from China Renaissance. Please ask your question.
Daniel, I have two questions. The first one on the-
Hi
balance sheet. Yeah, hi. It's Daniel. Last quarter, actually, you raised roughly $220 million through bank borrowing. But I noticed that it's a short-term debt. What's the reason behind? Is it because it's an evergreen loan that's subject to review every year?
Overall, for this year, for the past quarter, we borrowed just overall slightly. We have a $700 million loan facility. The $2.5 billion, $1.8 billion was equity. We own 51%. The IC fund owns 29%, the local government owns 20%, so that is the $1.8 billion. Then we have a syndicated loan facility by five banks, basically, in total, about $700 million.
Okay.
We are working on that loan. It is going to be signed off, I think, sometime this month. We need the money, so we basically borrow a bridge loan. There is no issue with that. We withdraw about $200+ million , maybe slightly under $240 million. It is a combination of it is a bridge loan plus working capital loan.
Oh, okay.
That was that. For the year, we are going to spend more money now. For the year, I think in Q4, overall, I think this year, we are probably going to be withdrawing about $500 million loan. It is a long-term debt, and then another $70 million on working capital. That is the expectation.
Oh, okay. Got you. Very clear. Yep.
Our overall debt ratio is still very low.
Yep.
Debt ratio is still very low.
True. Yeah. I just wonder why it is put under the short-term debt. Okay. Second question on the 8-in performance last quarter. Actually, it is pretty strong, right, in terms of the top line growth. But the gross margin was still kind of flattish, right, compared with Q2. What is the reason behind it?
Well, you are talking about specifically the 8-in business, huh?
Right. Mm-hmm.
Yeah, it was very strong. It was good. I think we had a very good. The ASP did increase about 2.5% overall. We didn't go out to lift the ASP, I think. That's not something we do because all our products, we have a market price. We basically always sell our products according to our pricing guideline. This is what our marketing salespeople do. Then, now the gross margin go up largely because the depreciation expense went up a little bit, and also the labor cost went up slightly because we pay the six-month, half-year bonus. The labor cost slightly went up, and plus depreciation expense went up. Just as I said earlier, look, you had a question on capital investment. We continue to upgrade the 8-in fab, so it actually went up slightly, about $2 million overall.
That probably explains why basically the gross margin didn't increase significantly. Or maybe was sort of compared to Q2, it came down a little bit. You know what? These things move so fast. I would expect our margin should be pretty stable. If the demand continues to be very strong, we have a very good chance to gradually improve our ASP. Okay?
I would expect in long term, I think our gross margin for the 8-in business will improve over time.
Okay. Very clear. Okay, thank you, Daniel.
Thank you. Your next question comes from the line of Jian Kuai from Orient Securities. Please ask your question.
Hi. Thanks for taking my question. The first one is regarding the product breakdown. We can see the revenue from logic and RF products increased to 49% year-over-year, and from communications increased to 48% year-over-year. Could you maybe give us more color about the huge increase from these products?
Well, a few things. I think for us, consumer is a very major important market. Industry and automotive have gone up very fast because of power discrete recently. Communications and some other segment, the reason it went up, you are talking about quarter to quarter, right? I think largely because we have this logic product. CIS is treated as under, it is a logic product, okay? That is why you see logic has gone up very fast. CIS is virtually important revenue driver for now at this point in our 12-in fab. I expect it will continue to be a revenue driver for the next few years. Okay? So that explains why it has gone up recently.
Okay, got it. So it is also the driver for the communication?
The driver for communications.
Okay.
Is that what you are saying? But that also has something to do with-
Yeah. CIS, right.
Yep.
Okay, got it. The second question is regarding the growth margin. You have already given guidance of 21%-23% in Q4. Maybe can you give us a breakdown of 8-in and 12-in? Just now, you mentioned the 8-in growth margin will continuously improve, so maybe give us a guidance in Q4.
Sure. I can do that. What I said, for the 8-in business with 27.2% in Q3, but I think we should be able to do better next quarter. Despite it's Q4, it's a year-end. There are a lot of elements into it, including bonus, but I think we should be able to have a better growth margin than the prior quarter. I would say it should be somewhere close to 28%. But the Wuxi fab, the depreciation expense will continue to go up. So overall, the growth margin will probably even be lower for Wuxi. So I think that was the primary reason why overall, we're talking about in the range of 21%-23%. But you know what? We'll do everything we can. This company, if you look at our results for Q3, we exceed our guidance on both revenue and growth margin.
Overall, we plan conservatively. We're hoping we can do better in the end.
Okay, got it. Thank you.
Thank you. Your next question comes from the line of Kenji Kawase from Nikkei. Please ask your question.
Oh, hello. Can you hear me?
Yes.
Well, thank you very much. I am Kenji from Nikkei. I have two questions. One with regard to U.S. sanctions onto your rival, SMIC. What sort of impact or benefit have you had since the sanction was imposed upon your largest rival? Also, a part of this first question, United States has a new President being elected. Do you think there will be any change in your environment doing business in the tech field with regard to the United States in the upcoming future? That is my first question.
Let me just answer your first question. We work with new customers on a daily basis.
Virtually, we have new customers knocking on our door weekly. We always work with our open arm. I cannot specifically answer you that question, but as long as there is a technology fit, we work with whoever the customer is. I can tell you that we did get a lot of new traction lately, from both new and existing customers. For the 12-in fab, our CIS product is a new customer. Our NOR flash is also a new customer, and we have new customers for other platforms as well for our new 12-in fab. But we also have a lot of existing customers also working with us on new products. We look at customers as a whole. We do not really look at where they come from, whether it is TSMC or a neighbor or somewhere else. But we are looking at the customers as a whole.
As long as we can serve them, as long as we can provide a better solution, as long as they accept that, we will work with them.
As far as the other question, I think it is still too early to tell. Can you perhaps rephrase that question very quickly again, just on the U.S.?
Yeah. I just wanted to know that the U.S., United States will have a new president, and how since under Donald Trump, the U.S.-China tension, especially in the tech area, has intensified, has been very stiff. I was wondering how you feel the overall environment situation is going to change or not under the new President.
Well, I think it is really too early to tell at this point whether there will be any changes with regards to export control. Hua Hong Semiconductor, we have always been in compliance with the rules and regulations on export control. We are one of the few VEU holders, validated end user. We have a very thorough and strict internal control process, including customer screening and good screening in place. From that perspective, I do not believe it affect us, or it will affect us.
Okay. Thank you. My second question is on the government subsidy. Looking at the number, it was $9.217 million that you received during the third quarter, which is almost double compared to the year before, and like 1,900% more than a quarter before. Would you please break it down on what subsidies they were? Were they from the central government or from the Shanghai government? Were they a much lower level government? Also, what was the title? Why did you receive so much more during the third quarter?
What is the number you are talking about now? Nine what?
Your government subsidy. Government subsidy under your broken down other income. It says that you have received $9.217 million in the third quarter, which is almost double, twice as much compared to the year before.
Yeah. Some of that it's for our new fab in Wuxi. Mostly, it was a tax credit, okay?
Tax credit.
For that sub-deal we negotiate with local government. Most of that. Also part of the other, it's a combination from two areas. One is most of that come from Wuxi and the other small part is from Shanghai. But it's mostly a tax credit, which we negotiate with the local government, when we decide to build a plant there. That is very important that we get local government support, not just in terms of policy, but in terms of the actual cash. So that is something we got, okay? It's very important to get this sort of support, especially the early stage of a fab, so that you want to have negotiate a good deal. This is not just China, this is global. Wherever you try to set up a fab. Even in the U.S., you would get credit as well. Okay?
It doesn't go forward then, since you will be continuing on with your investment in Wuxi, and therefore, the level of government subsidy is going to stay at the quite a high level then, the net upcoming quarter and-
If you are familiar with our company, like other analysts, we have told people that we have negotiate a fixed amount of subsidies that would help us in the early years, okay? It is a fixed number. I never publicly talk about this number, but we would get it basically in the first six years. Each year, we get a part of this overall amount. It would be more initially because of ramping up the depreciation expenses. It really meant to help us to offset some of that stuff. Okay? That is just in general, we have this kind of a subsidy agreement we have signed with the local government. But on top of that, there are other things we get from here and there. For example, I said there is a tax credit. Sometimes there are some projects that we work out that they give us some benefits. That is what they are.
Okay. Thank you very much.
Thank you.
Thank you. Your next question comes from the line of Chen Yaqing from Jiuren Capital. Please ask your question.
Thanks for the good question. My question is, can you share some more specific plans for the capacity ramp ups for Wuxi 12-in tomorrow? I understand it is 40,000 by the end of next year, but can you share more details about the quarter-on-quarter capacity ramp up and how we could do to develop our yield and when is our optimal break-even point for the Wuxi fab? This is my first question.
Okay. Basically, I answered all these questions in the past, but let me just be very brief on that again. We have about 20,000 wafer capacity. We expect to be in 20,000 in production or shipment by end of this year. That is the plan. We are already at the 15,000. We are running about 15,000 wafers in our fab right now, a month. We hopefully get to 20,000 by end of this year. We expect that will happen. At the same time, we are also expanding the capacity from 20,000 to 40,000. We hope we will get 40,000 capacity by end of this year, early next year. With the additional 20,000, we expect we have 20,000 by end of year.
We get to 25,000 by end of second quarter. We have 35,000 by end of third quarter, the supply. We will get 40,000 by end of next year. This is the manufacturing plan. As far as, this is the latest planning that I am aware of. This is how we plan to proceed. On the break-even point, as I said, we are hoping we can achieve by hopefully end of next year.
Got it. Thank you. My next question is, what are your plans for fab five and fab six? Do we have some additional capacity expansion plan or will it-
No.
Get into our Okay.
Well, we do not have a plan yet, but, for this fab, we have to first fill this fab first. As I said, the 40,000 will only fill half of the fab space. The next phase to make sure we fill the other half of the space, that will take another year or two. Okay? But certainly, as long as the demand and we have the technology to offer to the customers, I think we will, in the future, build another new fab, but that will be sometime down the road, not at this moment. We do not have a plan yet.
Okay. Thank you.
Thank you. Your next question comes from the line of Yiyang Luo from Founder Securities. Please ask your question.
Hi, Daniel. Thank you for taking my question. First question is about the Wuxi fab ASP. We just mentioned a lot of products of Wuxi fab, including smart card, DMOS, NOR flash and IGBT. Can you give some color about the fabrication fee of this product? Which is above the ASP and which is under the ASP, and how can we expect the ASP trend of Wuxi fab? The second question is about the competition. Power discrete seems to be a booming market next three years. Some fab like Shaoxing Zhongxin ramping up recently. How can we maintain our current competitive position? Thank you.
For your first question, we're running around $ 1,000 for the 12-in fab. At this point, we're talking about these are high volume stuff, not necessarily the most lucrative product, okay? They have high volume, but not necessarily the high margin. I think early stage, still early stage. We have to get the fab filled first. We expect, in the future, when we start to do embedded non-volatile memory, BCD and NOR flash, I think, the product mix should improve over time. Therefore, our ASP will also go up, okay? I would hope that, within a year, we should be able to get to $ 1,100, $ 1,200. That would be a nice, I would say, trend. Okay? The other thing is on your second question. What was the second question?
The competition.
Yeah.
Competition about the Shaoxing Zhongxin .
On that one, we're not concerned at all. Why? Hua Hong Semiconductor has the biggest foundry, discrete capacity as a foundry, okay? We've been doing this for so many years. We start with the general MOSFET, low voltage MOSFET business. We've been doing this for a long time, okay? We are as superior to any other foundry in power discrete. We're even better than, I think, the like of TSMC, okay? We're not concerned. There's a lot of demand for semiconductor. But to, A, be able to do what we're doing, for example, the IGBT superjunction, you need to take time. You probably will have to do it for a long time in order to get to that stage. Okay? It's a big market. Most of that high-end stuff, the high voltage stuff were imported from U.S., Europe in the past.
IGBT was virtually nil in China, okay? We expect this thing will continue to have great demand and there are more and more design houses are doing it. For IGBT, we're probably the only supplier right now, okay? Nobody else can get into this business because we're so good at it. It takes a long time for other people to be able to develop a good product. You know what? Whether it's international customers or if it's a local customer or local Chinese customers, they all know us. When it comes to power discrete, they come to us. Okay?
Okay. Very clear. Thank you.
Thank you. Your last question comes from the line of Sunny Lin from UBS. Please ask your question.
Hi, Daniel. Can you hear me?
Hi, Sunny. How are you?
Hi, I'm doing well. Thank you for taking my question. Sorry, I know it's a bit late, I will try to be quick. My first question is related to the question from the last person. It's on the ASP for Wuxi fab. I remember last time you mentioned that by maybe second quarter of next year, you expect ASP to get to maybe $1,500. But now it sounds like the ASP improvement will be more gradual. Is my understanding correct?
It is the product mix. We are doing many things there. If you look at our 8-in fab, because the power discrete in general, it is about $350 average. They go from $250 to always to about $500 or more for like superjunction. So the mix is about $350 - $400. Then you have the things like embedded. It goes anywhere $500 for the smart card ICs to $700 plus for the MCUs. So $1,500 is definitely achievable for similar products we are doing. Then the power discrete, if you take roughly $350 times 2.25, it would be about maybe $700 +. So it is going to be somewhere in between that range. But for some of the stuff we are going to be doing out, at this point, still a lot of that is 90 nm.
But once we get to 70 nm, 55 nm technology, $1,500 is a very achievable price for some of the products. They are all depending on the product mix that we are going to be. Things are changing, but at least they are changing for the better. We are ramping up very quickly. I know all of you cannot wait for us to get to $1,200, $1,300, eventually to $1,500. But we are certainly, just like everybody else, we are working very hard. We expect Q4 will be great quarter for Wuxi. You heard what I said. We are looking to expand the capacity there very quickly. So we have a lot of demand that we are seeing. We are working on a lot of products with some very high-rated customers, both in China and globally. So just work with us, be patient. We are moving very quickly on this. Focus on the revenue.
Do not think margin too much. Think about the good things like EBITDA. Margin, it is going to get better and better as we ramp up the fab, or the capacity.
Got it. Thank you for the color. That is helpful. My second question is on the depreciation for 12-in fab. What is the depreciation period for this fab? I would suppose that going to 2021 as you ramp to full capacity, potentially you would need to depreciate on 40,000 with a per month capacity basis. But based on the number you just provided, seems a bit low if I divide $2.5 billion by six. I am not sure if that is the right way to think about it.
You are on track. You are doing very well. This is how you do it, because you have to remember, we ramped from 20,000 to 40,000. We are not on a full 40,000 year. If we do a 40,000 in full, it would get close to $250 million - $300 million on depreciation just for Wuxi. Yeah?
Got it.
You see what I am saying? It is an average number that we are talking about. I would expect $170 million next year. That is what I said earlier, I think. It is still get to 40,000 by end of next year. It is mostly going to be average, and maybe it is around 20,000, 30,000 25,000- 30,000. So the depreciation is not in its full scale. But it is a year after, I think it will get to that number.
Great. Thank you very much for the color. Thank you.
Thank you.
Thank you. Ladies and gentlemen, that is all the time we have for questions. I will now hand back to Mr. Daniel Wang for closing remarks.
Well, again, thank you all for joining us today. Thank you for your support and the lots of enthusiasm. I look forward to talk to you again next quarter. Please continue to stay safe and healthy. Thank you.
Thank you. Ladies and gentlemen, thank you for your attendance. You may now all disconnect.