AXT Inc (AXTI) 2018 Q2 法說會逐字稿

Good afternoon, everyone, and welcome to AXT's Second Quarter 2018 Financial Conference Call.

Leading the call today is Dr. Morris Young, Chief Executive Officer; and Gary Fischer, Chief Financial Officer.

My name is Ashley, and I will be your coordinator today. (Operator Instructions) As a reminder, this conference call is being recorded.

I would now like to turn the call over to Leslie Green, Investor Relations for AXT.

Thank you, Ashley, and good afternoon, everyone. Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, we will provide guidance or make other forward-looking statements regarding, among other things, the future financial performance of the company and our ability to control costs, improve efficiency, increase orders in succeeding quarters, improve our competitive position in the market, our schedule and timeliness regarding our relocation plan, our thoughts on air pollution in Beijing, our ability to meet market demands for our products as well as other market conditions and trends, including those expected growth in the markets that we serve.

We wish to caution you that such statements deal with future events and are based on management's current expectations and are subject to risks and uncertainties that could cause actual results or -- actual events or results to differ materially.

These uncertainties and risks include, but are not limited to, overall conditions in the markets in which the company competes, global financial conditions and uncertainties, potential tariffs and trade restrictions, increased environmental regulations in China, market acceptance and demand for the company's products and the impact of delays by our customers on the timing of sales of products.

In addition to the factors that may be discussed in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission and available online by link from our website for additional information on Risk Factors that could cause actual results to differ materially from our current expectations.

This conference call will be available on our website at axt.com through July 25, 2019.

Also, before we begin I want to note that shortly following the close of market today, we issued a press release reporting financial results for the second quarter.

This information is available on the Investor Relations portion of our website at axt.com.

I would now like to turn the call over to Gary Fischer for a review of our second quarter results. Gary?

Thank you, Leslie, and good afternoon, everyone.

Total revenue for the second quarter of 2018 was $27.1 million. This compares with $24.4 million in the first quarter of 2018 and $23.6 million for the second quarter of 2017.

This represents year-on-year growth of 15%. Our total revenue substrate sales were $21.6 million compared with $19.4 million in the prior quarter.

Revenue from our raw material joint ventures was $5.5 million in Q2 compared with $5.1 million in Q1.

In the second quarter of 2018, revenue from North America was 8%, Asia Pacific was 67%, and Europe was 25%.

Also in the second quarter, no customer reached 10% of revenue, and the top 5 customers generated approximately 33% of total revenue.

We think this is noteworthy as it shows a growth that is diversified in both products and customers.

Gross margin in the second quarter was 40.6%, 4-0-point-6, compared with 39.2% in the prior quarter.

We are, of course, pleased with the gross margin, and let me give some explanation and color on this.

First, we had favorable product mix. Second, we benefited from an increase in the price of raw materials, which positively impacted margins at the 3 consolidated raw material joint-venture companies.

A third and smaller factor in Q2 is that we had previously written down gallium at one of the 3 consolidated companies, and so when they sell that material, the gross profit dollars are better. About 1% of our gross margin can be contributed to this in Q2.

Total operating expenses in Q2 were $6.5 million compared with $5.6 million in Q1.

Three items stand out that pushed up OpEx.

A onetime bonus payout at one of the subsidiaries of about $350,000, site consulting fees regarding the relocation of $115,000 and filing fees for business licenses and registrations as a result of the relocation of $86,000.

These 3 items totaled $555,000, of which at least the $350,000 is nonrecurring.

For the first half of 2018, our operating expenses are $12.1 million, or an average of approximately $6 million per quarter, which is in line with our run rate expectations for the year.

Total stock comp expense for the second quarter of 2018 was $464,000.

Operating profit for the second quarter of 2018 was $4.5 million compared with $3.9 million in the previous quarter and compared to $2.3 million for Q2 of 2017.

Interest and other income for the second quarter was a net gain of $400,000. This number consists of 3 categories: one, net interest earned of $200,000; two, accounting on our -- for our unconsolidated joint-venture companies, a gain of $300,000; and three, foreign exchange loss of $100,000.

The tax provision for the second quarter was $367,000 compared to $334,000 in Q1.

For Q2 2018, we had a net profit of $3.9 million or $0.10 per share.

By comparison, we had a net profit of $2.9 million or $0.07 per share in the first quarter of 2018 and $1.9 million or $0.05 per share in Q2 2017.

The diluted share count in Q2 was 40.216 million shares.

Cash, cash equivalents and investments closed at $54 million as of June 30. By comparison, at March 31, it was $67 million. The primary reasons for the decline were the new facility and equipment as well as an increase in inventory.

Depreciation and amortization in the second quarter was $1.2 million and capital expenditures were $8.7 million.

Accounts receivables net of reserves were $22.4 million at June 30, 2018, compared to $21.3 million in March of '18.

Net inventory at June 30 was $57.0 million compared with $51.1 million in inventory at March 31.

Ending inventory consisted of approximately 52% in raw materials, 44% in work in progress and only 4% in finished goods.

The increase in inventory was a deliberate decision based on 2 primary factors: first, with the ramp of our new manufacturing facilities, we are carrying inventory at multiple locations; second, the prices of both raw gallium and raw germanium have been increasing meaningfully, and our supply chain investments allow us to be a first responder to such changes.

As such, we have taken certain opportunities to purchase ahead in order to achieve a better cost structure for our substrate business.

It is important to note also that 80% of our total inventory consists of raw materials and ingot WIP.

At this stage, there is no customization and no shelf-life concerns.

This makes the possibility of obsolescence of notable concern, and additional 16% is wafer WIP, which can contain some customization.

A huge percentage is for existing high volume customers.

Finished goods is only 4%.

Looking ahead, we expect to be able to turn the total inventory number down over time.

Okay, this concludes the financial summary. I'll now turn the call over to Dr. Morris Young for a review of the business. Morris?

Thank you, Gary, and good afternoon, everybody.

Q2 was a solid quarter for AXT. We achieved revenue above our expectation as a result of a healthy demand for our products in every major category.

In addition, with a positive revenue mix and strong manufacturing execution, we posted one of the highest gross margins in our company's recent history.

For investors who have been tracking the performance of our raw material business, you will be pleased to see that in Q2, we posted a net gain from the 7 partially owned companies in the AXT supply chain.

In total, these factors contributed to our strong profitability in Q2, which more than doubled from the same quarter in the prior year.

We're achieving these results with the continued growth and emergence of a broad base of applications that require our products.

And there are others, such as 3D sensing, they are still to come, presenting a path of opportunity over the next 18 months.

As such, we are actively preparing our substrate business for growth in a number of ways, including the ongoing relocation of our gallium arsenide and germanium manufacturing, which is progressing according to plan.

Now for a review of our markets. Let me begin with indium phosphide.

We achieved another record quarter for indium phosphide revenue in Q2. Demand was solid from all the primary applications for our substrates. The power market was the largest contributor to our increasing sales this quarter.

Demand was particularly strong in China, driven by increasing number of broadband subscribers. We believe that the ongoing need for faster networks and increasing fiber-to-the-home requirements will continue to fuel this application though it may be a somewhat lumpy in the second half.

Meanwhile, we're beginning to see an improvement in demand for indium phosphide for data center connectivity, which are weaker in Q1 and Q2.

We believe that the continued adoption of silicon photonics technology in hyperscale and enterprise data centers as well as the transitioned over time to 100G and 400G technologies will drive the need for indium phosphide for many years to come.

These advanced technology increase capacity for transmitting data across many applications, reduce operation costs, allow for more scalable and simpler to operate systems and deliver significant reduction in footprint and energy consumption.

Now turning to gallium arsenide. Revenue from both semi-insulating and semiconducting gallium arsenide grew in Q2, primarily as a result of a wide variety of applications showing incremental growth.

Gallium arsenide appears to be experiencing a resurgence of demand, with new applications and advancement in existing applications driving its use.

The largest contributor to our revenue continue to be power amplifiers and Wi-Fi chips for wireless devices, LED lighting, signage and display and infrared applications. But we have seen smaller but meaningful demand from many other applications, including solar panels, surveillance cameras, biometric centers, horticulture lighting and more.

And going forward, emerging applications such as 3D sensing, lidar for autonomous cars, drone-based Internet connectivity and others are expected to provide another wave of growth.

We believe that the timing of these opportunities coincides nicely with our relocation and capacity expansion. We continue to make solid progress in bringing up new product -- production lines, hiring and training personnel and working with customers to fulfill their qualification requirements.

Our strategy is to execute the move in a measured and incremental way. This enables us to mitigate risks and provide a seamless transition for our customers while ramping up to meet our increasing demand.

By the end of the year, we expect to have relocated approximately 60% of our wafer production and expect it to complete the process by mid-2019.

In terms of qualification of new facilities, we are also making good progress. Our internal qualification results, to date, demonstrate consistent specifications across our sites.

In addition, we have completed a number of customer qualifications and are in the process with many more, including all of our major customers.

In total, we're executing according to plan and are encouraged by the results to date.

Now turning to germanium substrates. This segment of our business was, again, strong in Q2 as the (inaudible) industry continue its positive trend.

We typically see a fairly consistent demand for germanium substrate used in satellite solar cells, but since Q1 of 2017, our germanium substrate sales have grown by more than 70%, albeit from a smaller base of revenue.

Though we're not expecting that level of growth to continue, we do believe that positive market conditions are likely to provide an opportunity for sustained growth over the coming quarters.

In raw material, prices is now increasing meaningfully in Q2 and contributed to, both higher revenue from the 3 companies that we consolidate our financial statement as well as the improved profitability for the 7 companies we -- we account for using the equity method.

Q2 was the first time that these 7 companies had represented a collective gain to our financial statement in 10 quarters.

We're pleased to see the improved health of all our joint-venture companies as each provides strategic advantage for our business. As we have discussed through the years, not only have these companies been substantially accretive to our financial results over the history of our investment, they give AXT certain volume and cost advantages, particularly when raw material prices rise and capacity become more limited.

In addition, our partial ownership gives us early visibility into the changes in raw material market price, and this enable us to move quickly to take advantage of this knowledge.

It is important to note, however, that the benefit to rising raw material prices for our joint ventures is offset somewhat on the substrate side of our business by higher cost of goods sold. But on balance, this unique vertical integration provides tremendous value in managing the market dynamics for the materials that would -- that are critical to our manufacturing.

In closing, Q2 marked 20 years of us as a public company. Through all that time, I've seen AXT grow, change, and successfully adapt to new technologies, new applications and a dynamic market environment. It is to the credit of AXT employees, both past and present, that we have accomplished all that we have. After 20 years, AXT remains relevant to the technology that will shape the future, many of which are still to come. I am as excited as I have ever been about opportunities that lie ahead of AXT.

Our results this quarter and the way we position ourselves for a successful future provide further confidence in our potential and in the continued value we offer to our customers and shareholders.

This concludes my prepared comments, I will now turn the call back to Gary for your third quarter guidance.

Thank you, Morris. As we discussed, we are encouraged to see upside potential across the portfolio and believe we are well positioned in many applications that will drive our business growth in 2018 and beyond.

As such, we expect to see revenue in Q3 between $27.5 million to $28.5 million.

We believe our profit per share in Q3 will be in the range of $0.08 to $0.10 based on 40.316 million diluted common shares outstanding.

This concludes our prepared comments. Morris and I will be glad to answer your questions now. Operator?

(Operator Instructions) And our first question comes from the line of Edwin Mok with Needham.

My first question is on indium phosphide. Morris, you mentioned on the prepared remarks that you see some volatility around the PON market in the second half of this year, but you expect certain photonics to come back -- grow stronger. Are those offsetting -- in fact, you expect to see more growth in the coming quarters? Just kind of directionally, is that a way to think about that?

We do expect indium phosphide to continue to grow, but of course, the driver -- growth driver for the first half was mostly PON. But we think the silicon photonics probably was thought to shoulder more growth in the second half.

I see. So you expect overall to grow, but the driver will be different in the second half versus the first half. That's the way we should think about that, correct?

Correct.

Yes. Okay. And then on raw material, you said the prices has gone up quite a bit, but we haven't seen the -- the raw material revenue didn't really grow much. Is it because you're using more of that for your substrate, and therefore, even though your subsidiary are enjoying this higher price, you're not -- you're selling actually lower volume to your market? Is that how we should think about that?

Gary, any comments on that?

Well, it was up 10% Q-on-Q. So, no, I don't think there is anything, any real color or something to read between the lines. It's up. It's known in the public that the raw material's prices generally are up, and it's to be determined what happens quarter-to-quarter.

But so let me give you the color about joint ventures. We have 3 consolidated company. I mean, one is dealing with raw gallium, and the other is refined gallium and third one is pBN, we call it. But I think through the last 3, 4 years, the main contributor to our joint revenue has shifted more towards the pBN layer as well as refined gallium. The raw gallium that used to be the predominant revenue driver for our business -- joint venture business has now become a distant third.

Yes. And pBN by the way is for high-temperature crucibles, so we use those in our own ingot growth process, and the company that makes them also sells them in the open market.

And the other things -- I may comment is that although the raw material price has recovered, and so we turn from negative to positive, but nevertheless, it has not reached the all-time history. The gallium price are still selling for somewhere around, let's say, $200 a kilogram as compared to last year was only $110, $120 a kilogram, which is all-time low. But remember, back in 2011, gallium was selling for $800 to $900 a kilogram. So it hasn't reached that level yet. If it did, then our joint venture revenue would have been $10 million, I'm sure.

Great. That's certainly a helpful color. And then I just want a guidance, you guys are guiding for higher revenue, but the midpoint or EPS is lower. Is it just because you're modeling less comp profit contributing from the joint venture? I think that was a driver for the EPS for this quarter. Is that correct?

Yes. I think Edwin is saying we're guiding the revenue to be higher, but then the profitability, the EPS was not higher. So this may be for you, Gary?

Well, I think probably the gross margin had a little bit of help from this lower cost of market write-down that we've mentioned. And that's probably not going to be repeated the next quarter. So I think it's -- there might be some upside there, but I think it's probably in the range of $0.08 to $0.10.

I would also ask to say, no, you've got to compare guidance to guidance not the performance versus guidance. Because this quarter, after all, we beat by $0.02.

Right, right. Obviously, you did. Yes, it did really well this quarter. So I guess we should just stick with the mid-30s gross margin target for, kind of, how -- that's how you're targeting the model, and that's how you guys are modeling the business, right? Is this right like that?

Yes, I think I'm okay with somewhere in the 37.50%. So I mean that I was trying to keep the community at about 35% until we had a few more quarters to go by to make sure that this was real. But -- so I'd say mid-30s or maybe north of that towards 37%.

And our next question comes from the line of Gus Richard with Northland Capital.

Morris, can you talk a little bit about the qualification process? Maybe how many customers you have qualified and are they LED customers? Are they -- what sort of customers have qualified at this point?

Wow, Gus, that's a difficult question. I mean, I was looking at Excel files, table of the number of customer qualification sample we sent and the number of customers that we've sent to. I mean, it's -- I tell you, it's all across the board, all the way from LED to RF to germanium, and we're sending to qualification to a lot of people.

Can you give us the sense of how many?

It's all major customers. I would say hundreds of customers. I mean, whoever wants to qualify, and some of them doesn't really require qualification, but...

Yes. Universities and smaller, consistent players that order small quantities, they'll do, at best, a paper qual. They'll just look at our internal characterization data and compare. But the revenue-generating customers, virtually everyone has either -- has been given wafers, and some have finished and some are still in the process. The most important thing to note is it's pretty comprehensive. Secondly, our internal call data is pretty much virtually identical. I mean, that was the goal, that you could look at a wafer from the site, from the Tongmei site and compare it to the new site and not be able to tell the difference, and that's happening. We don't -- to be honest, Morris and I haven't sat and counted the Excel sheet. We've looked at it, but one of our senior level technical people is running the spreadsheet for us. I don't have an exact number, but it's very easy to say that all the customers that are large that have already been given wafers.

I guess, Gus, this question perhaps may be more pertaining to how many are we starting to ship from the new site. I think the answer to that is that our new site, although we're sending a lot of qualification sample to, but new site has not turned out to be a production facility yet. That's all now. And we probably expect that to happen in the next month or so, and beginning perhaps the first one to run real production is probably germanium wafers.

Got it. And then just on the gross margin, the upside from the previously written off material. How much did that add to gross margin in the quarter?

I think it added about 1%.

1%, got it.

And our next question comes from the line of Richard Shannon with Craig-Hallum.

I guess a question on the guidance for the third quarter on the revenues. Can you just give us a brief high-level thoughts on direction of growth of your major categories within the substrates as well as raw materials just to get a sense, relatively speaking? That would be great, please.

I think indium phosphide will continue to grow. Germanium is another growth driver. Raw material, we are figuring it's probably going to stay flat, and gallium arsenide, I don't think there is anything which is major. I think wireless is probably a little bit weak, but we have many other opportunities. One was the -- we're going to have some growth for the drone-based gallium arsenide solar cells. But you take gallium arsenide overall, I think it's a weaker growth than both germanium and indium phosphide.

It'll grow modestly in aggregate but it's probably stronger growth in the other 2 products.

Okay. Perfect. Maybe to dig in on one of those indium phosphide, it wasn't really clear to me the extent of contribution sequentially in the third quarter you're expecting from PON versus silicon photonics and other applications. If you can make sure we at least understand the direction, that would be great too.

Well, I think PONs market was really strong in first quarter and second quarter, but towards the end of second quarter, the PONs market started to turn south. There was a rumor in saying this customers have built some inventory and they -- but what they did say, will come back, but to what degree we don't know. But we do expect the silicon photonics to come back, but again, it's probably more back-end loaded towards the second half of third quarter. So I think it's like indium phosphide -- I mean, PONs market is a little bit worse to be -- than the last quarter, but silicon photonics start to come back up. But silicon photonics has one nice thing, is that if it comes back up, I expect the revenue to be much more pungent -- or powerful than PONs market.

And more stable.

Yes.

Okay. Into that topic on silicon photonics, do you get a sense of there is inventory being burned through here, or any other sense of what's going on with that? I believe that's more of a singular customer, but any color you could add there would be great.

Yes. That's why here, I mean, I think they have improved the yield, and some of the previously out of suspect for the matter of the thickness spec, it turns out that they can use it too. So they have a capability to expand their usage of our old inventory material, but which will burn through quickly, I believe. And then once they burn through the inventory, they will start to reorder.

Okay. Fair enough. Maybe one or two or more quick questions for me. I want to touch on raw materials, and based on a previous question that I may not have caught all the details here, I'd love to understand this study. You're talking about some improvements in raw materials prices, and I think you were saying that the raw gallium JV is actually a distant third in terms of contribution there. I've been tracking that gallium pricing for some time, and it actually appears to be kind of flattening out or maybe even rolling over here. But I think it is a raw gallium pricing number that I am tracking here. So is there a divergence in pricing between the raw and the refined gallium? And are the trends in the other items, I think the pBN crucible is, I think, is what you referred to, is that continuing to grow? Or if you can give us any sense of trends there, both revenue and profit-wise?

Okay. That's a mouthful of questions. So let me take that. The pBN, yes, I think it now turns out to be a very big contributor to our revenue for our JVs. And our pBN crucible company is called Boyu. They used to be doing pBN for gallium arsenide crucibles. Now they've found a new business line, which is doing OLED. OLED, there is a lot of manufacturing components which requires pBN's, high-temperature stability properties. So they are growing that business very nicely. As far as -- so they have grown both revenue as well as bettered profitability. They are expanding the capacity in a big way. So we do expect them to do even better in the future quarters. And the other thing that happened to this pBN crucible company is that they had a German competitor who liked their product so much, they came in and buy into a 10% shareholder in the last quarter, spending about $2 million to buy into our Boyu JV third partner. So as a result, you can see that, that joint venture is now valued at $20 million in equity value. So now, getting into raw gallium, the raw gallium, the low -- all-time low price was 700 RMB per kilogram. I've been tracking on that -- with renminbi, I'm sorry, I didn't exactly change it to U.S. dollars, but you can certainly divide it by 6.5 and get the U.S. dollars. That has increased to about RMB 1,500 per kilogram by sometime in Q2, but as you said, it's now rolling over back to around RMB 1,250 per kilogram. So although it's not as great as the recent peak, but certainly it's much better than what used to be, the bottom of the market, which is about RMB 700 to RMB 800 per kilogram. So did I answer most of your questions?

Yes, you got them all, Morris. I'll just throw a last one in for Gary. On the nonconsolidated joint ventures, you noted a positive contribution first time in 10 quarters. Do you expect that trend to continue, i.e., to remain profitable for the foreseeable future? Or is there a different trend we should think about?

Well, you know me, I like to see it happen a few quarters in a row before I stick my neck out. I think it's -- I'm not sure it's going to hit any home runs, but I think there is a high likelihood it'll be at least break even or a little bit north of that. So then we'll have to see if it really takes -- gets on a run or not.

I think the market situation to me, it's a little bit complicated because first of all, it is a trade issue and the world economy, it's nobody knows which direction it will go. But normally, when you see a raw material, it usually has a very long cycle. When it goes down, it goes down 5, 6, 7 years and when it goes up, it goes up at least 3 or 4 years. So now we're in the first beginning of the cycle, I think. So how long will it go, and this increasing demand for gallium is actually coming from more gallium arsenide, and I saw another article in Semiconductor Today which is predicting gallium arsenide to grow tremendously, fueled by this LED demand, the pixel demand. That always going to put pressure on the capacity of gallium production. So I don't think -- I mean, sure, it's going to go up and down a little bit, but it's not a straight direction. But I do think that the demand, it will start to increase, and that will increase the price of all this raw material.

And our next question comes from the line of Hamed Khorsand from BWS Financial.

So first question I had was, are you generating any revenue from the wafers that you're selling for the qualification process on the new facility?

Yes. Usually, I don't think it's free samples. I think it's, we do deliver our qualification wafers with -- we have to get paid.

Yes. It's not a meaningful number, Hamed. It's not going to be moving the needle in any big way, so.

Understood. And then are you managing the startup process of the new facility, given that you're seeing some maybe like sluggish growth, is probably the best way to say it, as far as what's going on right now? As far as anything capacity so there is no oversupply?

I don't think we said sluggish growth in any long-term way.

Well, for us...

I said the PON market was slowing down.

But that's determined by the market, the demand, not how much we can produce. So I think we have plenty of capacity.

He's asking, are we building too much?

Oh, are we building too much? I don't think so. I mean, we don't build -- I mean, we do have some redundant inventory, as we stated in our prepared comments, but it's all because there are customer demand, not -- we're not -- we don't build for nothing.

The footprint of the buildings will have some extra space. That's deliberate. That's not very expensive to include when you're building. And then as I've said before, the long pole in the tent for adding capacity is the facility. Once the facility is facilitized and in place, then we can get equipment moved in much more quickly. So facility might take a year, but equipment may be, from decision point to getting it in the line and tuning it up, could be 90 days. So the strategy has been to make sure we have the facilitization and the footprint available and then we can add furnaces pretty fast.

And just given that you're still in this transition of moving to the new facility, how sustainable are gross margin at these levels? It's been quite a bit of outperformance to what I was expecting.

Well as I said, I think the gross margin, a reasonable place to be thinking is in approximately 37.5%. And I don't think it's so much necessarily because where the facilities are happening, but as the overall price of raw materials goes up, we benefit below the operating line on the 7 companies because they make more money, but in general, we will pay more for raw materials, especially raw materials that are not bought from the 3 consolidated joint ventures because the accounting there is different. So that's going to be a bit of a drag. I don't -- no one should be forecasting north of 40%, and we expect it's going to drift down a little bit, but not because of any single thing about the facility. It's more about the influence of raw materials.

Yes. But there's also these redundant and fixed costs though, and so far you've put up 2 quarters in a row above 37.5%, and if I include Q4, it was 37.2%. So is it more about leverage in the operating business? Or is it all just the product mix?

No. There's the overall increased volume, then it absorbs more deeply, if you will. So you're -- we sometimes can be under absorbed if the volume's, let's say, south of $25 million. And when we're in these ranges, then the absorption of the fixed cost is better.

But I think Hamed may have a point, is that sometime in the future, let's say mid-2019 or early 2020 when our new facilities start to come online. Now you're going to start to put that into your fixed overhead and you're going to have dual location or 3 locations give you that burden to absorb. But then as we have always been saying, we then will evaluate our Beijing facility. We would start to decommission it, and...

Yes, Morris is correct. There will be some overlap, and if there is overlap and at the same time there is a revenue decrease and we start dipping south of $25 million, it's going to be noticed. But if the revenue continues to tick up a bit from quarter-to-quarter, it'll probably be not noticed; it'll just be absorbed in the increased volume.

This other way you calculate it was that the kind of facility we build out are depreciable in 25 years.

27.5.

27.5 years, so per quarter, you think we have double counting on those facility, using elimination is something like $0.5 million a quarter or something like that.

Or less.

You can sort of quantify that. I mean, I don't think it's going to be meaningful. I think, yes, you're right, if we could start to get -- I'm excited about the data center business in 2019 as well as pixel and LEDs, and then if revenues start to go north, then all those are...

We'll be glad we have all that equipment in the facility. So the -- we will eventually -- so there will be some extra depreciation expense, and the short answer is that we hope it will be absorbed by increasing revenue growth and volume growth. And then a side note, as Morris alluded to, is that at some point, we will decommission some of the utilizations in Tongmei, Beijing. And so then those, if they're not in service, in use, then the GAAP accounting rules would say that you would have to stop depreciating them and then they're subject to the impairment valuation tests, which would be is the site -- is the book value of the site or the piece of the element of the site that you're examining, let's say it's part of the building or something, is the book value of that item above the market value or below the market value? And so we will have to perform those tests. However, because the whole thing in the Beijing district that we're in is about the redevelopment of that section called Tongzhou. The real estate prices are -- have moved up pretty dramatically in the last 2 or 3 years and they'll probably continue to increase. So I don't -- I can't say for sure, but I think it's very, very unlikely that there'd be a facility impairment valuation.

And our next question comes from the line of Drew Burke with Game Plan Financial Advisors.

I just want to follow up on kind of long-term demand for gallium arsenide substrates. And specifically, there was a comment by Corvo at their analyst meeting with regards to thinking that gallium substrate would win back share in 5G cell phones that was lost, I believe, back in the '11, '12 timeframe. So are you guys seeing any customer activity along those lines? Do you believe that to be the case? And could you quantify -- I mean, I know it was a big deal in AXTI's history when that business went away in 2011. I know it was a big deal in gallium prices shortly thereafter when that volume was lost. Can you guys comment on any of that?

Sure. So this is a very interesting question. Yes, you correctly reminded me how painful that was when gallium arsenide lost to SOI for the PON business for the switching business for the cell phones. But I think looking ahead for the 5G, I think most of the markets that we think that 5G will use higher frequencies, the first step is going to 6 gigahertz and the next step is going to 28 gigahertz. Now for 6 gigahertz, I believe that gallium arsenide probably can fit that market very well, but the number of chips will be used, I don't believe it's going to dramatically increase the usage of gallium arsenide. Okay? But I will say the opportunity for AXT looking at 5G, it's rather several-fold. One is, when you go to 28 gigahertz then the speed of those handheld devices probably will start to require something more sophisticated and much more mature property-wise, probably will start to require indium phosphide used as a call amplifier. That will be a very good benefit to our indium phosphide substrate business. But perhaps not a whole lot of big opportunity for gallium arsenide, although gallium arsenide, if you tune it, the device properly right, it probably can reach 28 gigahertz too. But I think indium phosphide, because of these natural property, is going to perform much better at much higher frequency. And the next frequency for 5G is actually 77 gigahertz. As you see those higher, much higher frequency, that requires a different material, and indium phosphide by nature can handle higher frequency at a much -- easily. Okay? So there is another opportunity for 5G. That is, when you [don't] 5G network, you will need a lot of cell phone towers. Right now, 4G, you have a cell phone repeater every square mile. And so you see those cell phone towers along the highway. But when you deal with 5G, you probably need, they call it, (inaudible) cells or micro cells. So every room, every building, everywhere you go, you will need a cell to act as a repeater, because higher frequency is very difficult to make it into a very high power as well as the higher frequency doesn't turn corners. So as a result, all those micro cells need to be connected. When you connect those micro cells, from what we understand, you need indium phosphide fiber to connect all these fiber, thousands and tens of thousands of micro cells into a central station. And now when all those central station will then go into a data center or switching center to take care of the signal processing, because every mobile phone and every autonomous car will generate 10 to 100 times more data, and that will increase the traffic for the data center to a hundredfold. So you can see in the analysis of Cisco, they expect that this data consumption to skyrocket by a factor of 100. And so the data center business, actually, I believe we're going to benefit from twofold. One is connecting all these micro cells together and feed it into the cell tower, and then the data center, we're processing hundreds more times the data. Plus you can also cannot have a delay, because if you're going to make a decision for that autonomous car to make a left hand turn or a right hand turn, you have to do it very fast. So the best way to processing those data in a lightspeed is silicon photonics, it's indium phosphide. So I don't know if I answered that question.

No. That's very helpful. I would ask for clarification, though. You did say in your prepared remarks that the gallium -- some of the gallium strength, a driver to that was power amplifiers. And I hadn't heard power amplifiers being a driver in your gallium business in quite some time, so that led to the questions. I was just seeking clarification, but thank you for your feedback.

Anything else, Drew?

And I'm not showing any further questions at this time. I would now like to turn the call back over to Dr. Morris Young for any further remarks.

Thank you for participating in our conference call. We will be participating in the Jefferies 2018 Semiconductor, Hardware & Communications Infrastructure Summit, Tuesday, August 28 in Chicago. And Dougherty and Company 2018, as you should know, investor conference on Thursday, September 6 in Minneapolis. We look forward to seeing many of you there. As always, please feel free to contact me, Gary Fischer or Leslie Green directly if you would like to meet with us. We look forward to speaking with you in the near future.

Thank you. Ladies and gentlemen, thank you for participating in today's conference. This does conclude today's program and you may all disconnect. Everyone, have a wonderful day.