Semtech Corp (SMTC) 2018 Q2 法說會逐字稿

Good afternoon. My name is Amanda, and I will be your conference operator today. At this time, I would like to welcome everyone to the Semtech Corporation Q2 FY '18 Earnings Release Conference Call. (Operator Instructions) Thank you. At this time, it is my honor to turn the call over to Mr. Sandy Harrison, Director of Finance and Investor Relations. You may begin your conference.

Thank you, Amanda, and welcome to the Semtech's conference call to discuss our financial results for the second quarter fiscal year 2018. Speakers for today's call will be Mohan Maheswaran, Semtech's President and Chief Executive Officer; and Emeka Chukwu, our Chief Financial Officer. A press release announcing our unaudited results was issued after the market closed today and is available on our website at semtech.com.

Today's call will include forward-looking statements that include risks and uncertainties that could cause actual results to differ materially from the results anticipated in these statements. For a more detailed discussion of these risks and uncertainties, please review the safe harbor statement included in today's press release as well as the other Risk Factors section of our most recent periodic reports filed with the Securities and Exchange Commission. As a reminder, comments made on today's call are current as of today only, and Semtech undertakes no obligation to update the information from this call should facts or circumstances change.

During the call, we will refer to non-GAAP financial measures that are not prepared in accordance with generally accepted accounting principles. Discussion of why the management team considers such non-GAAP financial measures useful, along with detailed reconciliations of such non-GAAP measures to the most comparable GAAP measures, are also included in today's press release. As a reminder, all references to financial results in Mohan's and Emeka's formal presentations on this call refer to non-GAAP measures unless otherwise noted.

With that, I will turn the call over to Semtech's Chief Financial Officer, Emeka Chukwu. Emeka?

Thank you, Sandy. Good afternoon, everyone. For Q2 fiscal 2018, GAAP net sales were $153.1 million, an increase of 6% sequentially and 13% year-over-year. Q2 GAAP net sales included $3.2 million of expense for the Comcast warrant. In Q3, we expect the charge for the Comcast warrant to increase to approximately $6.7 million as Comcast accelerates its rollout of network coverage. We now expect the Comcast warrant expense to come in around $23 million for fiscal year 2018. Q2 GAAP gross margin increased 110 basis points sequentially to 60.2% due to lower sequential Comcast warrant expense and a more favorable product mix.

Q2 GAAP operating expense increased approximately 10% sequentially due to the impact of higher stock price on equity compensation, higher new product expenses and transaction-related expenses. Q2 GAAP tax rate was 24.6% compared to 24.1% in Q1. For fiscal 2018, we expect our GAAP tax rate to be in the 25% to 29% range.

Moving on to the non-GAAP results, which exclude the impact of share-based compensation, amortization of acquired intangibles, acquisition and disposition related and other nonrecurring charges not tied to current operations. Q2 net sales were $156.3 million, a 5% sequential increase and 15% increase year-over-year and the seventh consecutive quarter of results above the midpoint of our guidance.

In Q2, shipment into Asia represented 73% of total sales. North America represented 19%, and Europe represented 8%. Total net sales to distribution represented approximately 65% and direct net sales represented approximately 35%. Q2 bookings slowed sequentially but increased year-over-year and resulted in a book to bill below 1. Those bookings accounted for approximately 37% of shipments during the quarter.

Q2 non-GAAP gross margin was 61.2%, an increase of 30 basis points sequentially due to more favorable mix. We expect our Q3 non-GAAP gross margin to be slightly higher sequentially due to improved manufacturing efficiencies.

Q2 non-GAAP operating expense was $52.9 million, up 3% from Q1 levels due to higher new product expenses. In Q3, we expect non-GAAP operating expense to decline slightly due to timing of new product expenses and lower variable compensation. We continue to expect our non-GAAP operating expenses for fiscal year 2018 to be flattish year-over-year. In Q2, our non-GAAP tax rate was 19.1%. We expect our fiscal 2018 non-GAAP tax rate to be between 18% and 20% based on current regional income assumptions.

In Q2, cash flow from operations increased 244% from Q1's typically weaker quarter to approximately $36 million. Cash and investments in Q2 were $278 million, and our debt balance at the end of Q2 was approximately $234 million. In Q2, approximately 73% of cash and investments were domiciled in international accounts and 27% was based in the U.S.

During the quarter, we acquired AptoVision, a provider of algorithms for transporting video over IP for approximately [$18 million] in cash, net of cash acquired and certain working capital adjustments and a contingent consideration of up to a maximum of $47 million over 3 years if certain goals are achieved in each of the earnout periods. We expect AptoVision to contribute approximately $2 million of revenue in fiscal 2018 and be neutral to our fiscal 2018 non-GAAP results and to contribute between $6 million and $10 million in revenue in fiscal 2019 and be modestly accretive to fiscal 2019 non-GAAP results.

In Q2, we made approximately $2 million of investments in startup companies focused on eliminating bottlenecks in the LoRa ecosystem. We repurchased approximately $400,000 or 12,000 shares of stock, and the outstanding stock repurchase authorization stands at approximately $51.4 million. The primary use of cash continues to be the pay down of debt.

Next, strategic investments and opportunistically repurchase our shares. In Q2, accounts receivable increased 9% sequentially due to higher net sales. Our days of sales outstanding increased a day to 34 days and remains below the target range of 40 to 45 days due to a higher mix of shipments to distributors.

Net inventory in absolute dollar terms decreased approximately 2% sequentially and represented 114 days of inventory and above the target range of 90 to 100 days. In Q3, we expect our inventories to be sequentially flat in both absolute dollars and days.

In summary, we are pleased with the strong first half results. We delivered our seventh consecutive quarter of strong performance as our differentiated growth drivers and diversification of strategy continued to drive solid results. Our expanding gross margins and OpEx control continue to deliver a strong operating leverage and should contribute to the record financial performance we are expecting in fiscal year 2018.

I will now hand the call over to Mohan.

Thank you, Emeka. Good afternoon, everyone. I will discuss our Q2 fiscal year 2018 performance by end market and by product group and then provide our outlook for Q3 of fiscal year 2018. In Q2 of fiscal year 2018, we achieved non-GAAP net revenue of $156.3 million, an increase of 5% sequentially and 15% over the same period a year ago. We posted non-GAAP gross margin of 61.2% and non-GAAP earnings per diluted share of $0.48. In Q2 of fiscal year 2018, demand from the enterprise computing market increased over the prior quarter and represented 34% of total net revenues. Demand from the high-end consumer market increased from the prior quarter and represented 29% of total net revenues. Approximately 23% of high-end consumer net revenue was attributable to handheld devices, and approximately 6% was attributable to other consumer systems. Demand from the industrial end market increased over the prior quarter and represented 25% of total revenues. Finally, demand from the communications market declined sequentially and represented 12% of total net revenues.

I will now discuss the performance of each of our product groups, beginning with our Signal Integrity Product Group. In Q2 of fiscal year 2018, net revenue from our Signal Integrity Product Group declined slightly from the prior quarter and represented 43% of total net revenues. During the quarter, our data center business once again delivered record quarterly results. Demand for our ClearEdge 100 gigabit per second CDR platform remains strong, driven by the increased demand for 100 gigabit per second links in cloud and mega data centers and the higher penetration of our CDRs in 100 gigabit per second modules. We recently announced our first 100 gigabit per second PAM4 PMD platform named [Fiber Edge], which includes our first 100 gigabit per second linear driver and our first 100 gigabit per second TIA. Our Fiber Edge platform is an excellent complement to both our ClearEdge CDR platform and our future MultiPhy single-lambda DSP platform. Fiber Edge opens up a new SAM for us in the 100 gigabit per second market, and we are seeing very strong interest from customers designing next-generation PAM4 solutions. Fiber Edge provides a seamless interface to PAM4 optics tailored for 100-gig, 200-gig and 400-gig modules.

Progress on our single-lambda 100 gigabit per second PAM4 DSP platform through our partnership with MultiPhy is also progressing well as MultiPhy is now sampling selected customers with its latest 100 gigabit per second silicon. We believe that the combination of the MultiPhy DSP platform together with our new Fiber Edge platform ideally position Semtech to take advantage of future 100 gigabit per second, 200 and 400 gigabit per second data center applications.

In Q2, our PON business declined. We expect the 10-gig PON market to grow and somewhat offset weakness in the 1-gig and 2.5-gig PON markets in future quarters. In Q2, our wireless base station business also declined. We anticipate the China wireless base station market to remain relatively soft until 5G deployments in China begin in calendar year 2019.

During the quarter, we acquired AptoVision, a leading supplier of advanced video transport algorithms used in the processing of video over IP applications. We estimate the current target SAM for the video over IP Pro AV market to be approximately $150 million today, and we expect this market to increase nicely over the next several years as video consumption and distribution becomes pervasive in numerous industrial applications such as medical, education, transport, retail and stadium displays and many other applications.

The combination of AptoVision's advanced video transport algorithms and Semtech's leadership technology position in ultra-high definition video broadcast positions us well to lead the video over IP Pro AV market, which we believe will become yet another future growth engine for us. With the steady pace of new products and the increasing SAM we are targeting, we believe our Signal Integrity Product Group will continue to deliver double-digit year-on-year revenue growth, which it has produced over the last 5 years. For Q3 of fiscal year 2018, we expect net revenues from our Signal Integrity Product Group to be down as continued strength in data centers is offset by ongoing weakness from China infrastructure.

Moving on to our Protection Product Group. In Q2 of fiscal year 2018, net revenue from our Protection Product Group increased 7% sequentially and 24% over the same period a year ago and represented 29% of total net revenues. Demand from our smartphone customers increased over the prior quarter, driven by the ramp-up of new flagship smartphone models. This strength was somewhat offset by further weakness in the China smartphone market.

While we continue to gain share in the mobile market due to the high quality and high performance of our products and the increasing need to protect more sensitive interfaces, we are also seeing increased demand for our protection devices in the automotive, IoT, high-speed computing and OLED display segments. While these segments typically have longer time to revenue cycles, the positive momentum serves to slightly offset the volatility of the handheld business.

During the quarter, we introduced several new products that address the unique protection needs of high-speed data ports. These included the RClamp0532T targeted at protecting dual-line 2.5-gigabit, 5-gigabit and 10-gigabit Ethernet lines used in data centers and telecommunications backplanes and the RClamp3321P targeted at high-speed data lines. We believe the customer diversification within our smartphone business along with the adoption of protection in new markets will continue to drive growth in our protection business over the next several years. In Q3 of fiscal year 2018, we expect our protection business to increase over the prior quarter on higher demand expected from our existing smartphone customers and initial penetration of North America's largest smartphone manufacturer.

Turning to our Wireless and Sensing Product Group. In Q2 of fiscal year 2018, net revenues from our Wireless and Sensing Product Group increased 19% sequentially and 59% over the same quarter a year ago and represented 21% of total net revenues. Q2 results represented a new quarterly revenue record for our Wireless and Sensing Product Group led by record revenues from both our LoRa platforms and our proximity sensing platforms.

Our LoRa business, as anticipated, achieved another revenue record this quarter as our position in the IoT market continues to strengthen. We see a rapid acceleration of LoRaWAN network deployments across the globe where end customers are either replacing older cellular networks or building complementary LoRaWAN networks for emerging IoT applications. Membership in the LoRa Alliance surpassed 500 members for the first time in Q2, including the addition of several industry-leading companies such as Alibaba and Charter Communications. And we are expecting the number of announced LoRa network trials or deployments to exceed 60 countries by year-end.

We are also anticipating that the number of deployed macrocell gateways will now reach 65,000 by the end of the year. This equates to a sensor capacity of approximately 300 million sensors. We anticipate this gateway footprint will continue to increase and ultimately enable connectivity of billions of LoRaWAN sensors within a few years.

Other important developments announced during the quarter included Comcast announced their decision to go from trials within 3 cities to full deployments in 15 cities as the next phase in build-out of their nationwide U.S. LoRaWAN deployment. In addition to Comcast's network deployment, it is starting to focus attention together with Semtech and other ecosystem partners on providing complete IoT solutions to end customers. Tata announced their decision to deploy LoRaWAN in 60 cities in India by the end of 2017 as they build on their smart city initiative across the country. NNNCo, a provider of nationwide narrowband networks announced plans to build a nationwide rural LoRaWAN IoT network to bring high-tech agriculture solutions to Australian farmers. The network will deploy sensors that use limited power and can operate in the field for years without the need for intervention. X-TELIA announced deployment of a LoRa IoT network in Québec to provide the necessary infrastructure to develop industries that require highly secure, low-power, long-range and low-cost connectivity.

While network deployments continue to accelerate, we are also seeing explosive growth in the availability of LoRaWAN sensors and end nodes being developed and announced, the broad range of sensors across many different end markets, including smartphones, smart cities, smart agriculture and smart health applications, to name a few. Some of the recent sensor announcements from our ecosystem partners include TrackNet announced its Tabs wristbands and object locators and other sensors using LoRa technology. TrackNet is using its LoRa-based Tabs solution as a way to bolster deep indoor network coverage and a cost-effective way to ensure high-quality network connectivity.

Movtek announced the development of a LoRa-based smart watch and smart bracelet that can track the user's location and can be worn by senior citizens and children who need monitoring for safety. And YoSmart announced they have developed a series of LoRa-based devices for smart home applications, including a smart lock, smart thermostat, smart sprinkler system and smart electric outlet. We continue to be very excited by the progress of LoRa across the globe as the number of low-power sensor networks and cloud-based IoT services starts to increase. The LPWAN market is clearly an exciting emerging space, and we believe LoRa is ideally suited to lead this market.

In Q2, our proximity sensing business also delivered a new quarterly revenue record as mobile device manufacturers continue to address increasing global concerns around excessive smartphone RF emissions. Many countries are adopting stricter RF emission standards for smartphones, which are driving increased interest and demand for our sensing platforms. For Q3 of fiscal year 2018, we expect net revenues from our Wireless and Sensing Product Group to be down modestly as we expect another record quarter for our LoRa Platforms to be offset by slightly lower proximity sensing demand.

Turning to our Power and High-Rel Product Group. In Q2 of fiscal year 2018, our Power and High Reliability Product Group rebounded from the seasonal weakness experienced in Q1 and increased 6% sequentially and represented 7% of total net revenues. Our Power and High-Reliability Product Group delivered another strong quarter of new product releases with 10 new products, including our latest -- the latest member of our FemtoSwitch family. This low-power load switch is our first automotive qualified load switch. We also released our first multi-system wireless charging platform for low-power wearables. This platform enables multiple wearables to be wirelessly charged at the same time and is targeted at smart watches, smart jewelry, smart hearing aids and other low-power systems. Design win activity for our wireless charging and switch products remains solid, and we expect industry adoption of wireless charging and relay replacement products to increase over the next 12 to 18 months. In Q3 of fiscal year 2017, we expect net revenues from our Power and High-Reliability Product Group to increase.

In Q2, the total company distribution POS increased and achieved a new quarterly record. Q2 distributor days of inventory was flat with Q1 at 70 days and remains at the lower end of our targeted range of 70 to 80 days. Our distributor POS business remains well balanced with 59% of the total POS coming from high-end consumer and enterprise computing end markets and 41% of total POS coming from the industrial and communications end markets.

Moving on to new products and design wins. In Q2 of fiscal year 2018, we released 20 new products and we achieved 2,069 new design wins.

Now let me discuss our outlook for the third quarter of FY 2018. Based on current bookings trends, we are estimating Q3 non-GAAP net revenues to be between $152 million and $160 million. To attain the midpoint of our guidance range, or approximately $156 million, we needed net turns orders of approximately 46% at the beginning of Q3. We expect our Q3 GAAP earnings to be between $0.18 and $0.22 per diluted share and our Q3 non-GAAP earnings to be between $0.48 and $0.52 per diluted share.

I will now hand the call back to the operator. And Sandy, Emeka and I will be happy to answer any questions. Operator?

(Operator Instructions) Your first question comes from the line of Cody Acree from Drexel Hamilton.

Mohan, if we could just kind of talk about the guidance for a bit. So you've been growing very nicely annually and sequentially. You're guiding out about flat to the midpoint. I guess, what areas maybe surprised you during the quarter and are leading to this flat guidance?

Yes. I think, Cody, the China -- situation with China PON base station and handsets is just a little bit weaker. Q3 is typically stronger for smartphones, and China smartphones is a little bit weak. And then what we're seeing on the PON side and infrastructure side in China is also a little bit weaker than we had anticipated. We thought it would -- some of it would come back in Q3, and it's just -- looks like it's going to be another few quarters before that picks back up again. So I guess, those are the main areas that we see a little bit softer than we anticipated. It's not a huge amount, but it was enough to surprise us a little bit.

Sure. And are you getting any visibility to the end of the softness? Is it a matter of Chinese inventory burn in the smartphone market or PON deployment projects? Just what is your level of visibility?

Yes. On the smartphone side, there are indications that Q4 will be -- will see a little bit of a pickup in the China handset market. I would say, in general, my comments on China smartphones are that they're still learning the business, and the share participation between the different manufacturers in China are still being worked out a little bit, so some of that. So inventory levels are perhaps a little bit higher in Q2 and Q3. But we'll start to, I think, improve in Q4. On the PON side, we see 10 gig is doing quite well, I think, in general, and we expect that to grow. But the 2.5 gig and below is -- seems to be weaker, should start to see a pickup in Q1 for sure, maybe a little bit pickup in Q4. But base station, China base station looks like it's going to be fairly soft at least for most of this year and we think some of the first half of next year for sure.

And then lastly, you briefly -- you've made a very brief mention of a new protection win with a major North American smartphone player that I don't think has necessarily been considered. So can you maybe talk about what's changed with that customer or that opportunity that had not been there in the past? And then how pervasive might this be for you?

Well, we've always said that we think our protection products are the best in the industry and that we have a unique differentiation for small form factor, high performance, high bandwidth kind of interfaces. And as the industry moves to more advanced lithographies, we've always said that market will come to us to some extent in those segments, and that's what's happening. And I think we have opportunities in all regions and all manufacturers, and we have some nice design wins that, I think, will start to ramp in Q3 and really, we'll start to see develop in Q4 next year. We're anticipating between a $3 million and $5 million quarter run rate if everything plays out well starting in Q4.

Your next question comes from the line of Harsh Kumar from Stephens.

So I think an appropriate topic is to follow up on the previous question with the large North American customer. What exactly, I guess, is the functionality of some of the parts, Mohan, if you can possibly address that?

Well, I can't give any details, obviously, Harsh, but it's the same protection products that we use in other smartphone -- with our other smartphone manufacturers. We protect interfaces, high-speed interfaces as you know, and it's really no different. Obviously, the smaller form factor of our Z-Platform is a critical differentiator for us. Also, the ability to protect against advanced lithography in those is also another key differentiator for us and then the high-speed interface also. So it's -- as I've said, we believe we are the leader in this space, and there's a reason why we are doing so well not only in handheld but anywhere where there's advanced lithographies and high-speed interfaces and a small form factor requirement. And so the market's going to come to us. And as I've said, if doesn't happen in this generation, it will happen in the next one. So it's going -- it's starting to -- obviously, our success in Korea and China has now been -- we've been very successful in those regions, and it's nice to see some progress in North America also.

If I can ask you on -- I noticed that your turns requirement for this current quarter is 46%. It seems to be a little bit higher than history. Maybe, Emeka, you could help me understand how I should think about this or what kind of level of visibility you have on those turns. I mean, they're turns but whatever color you can give me.

Well, Harsh, the only thing I can add is just that actually if you back to a year ago, the turns requirement for our Q3 was probably something in the same range. So I think this is just very seasonal -- is typical for us to have a higher level of turns in the quarter just because of the mix of revenues. In the third quarter of the year, we typically have a stronger consumer business, and the lead times over there are a little bit shorter than what we have in our other product...

Understood. Sorry to go back and forth. Mohan, I wanted to ask -- there is one -- at least one manufacturer of data center optical parts that talked about a transition or some issues going from 40 gig to 100 gig. They echoed your sentiment that 100 gig is strong, but the 40-gig piece was falling apart a little bit for them. I was curious if you could tell me or maybe qualitatively discuss how much exposure you have on the 40 versus the 100 and if this is an issue for you.

We don't -- I don't think we have that much exposure on the 40-gig side, Harsh. We do have on the 10-gig side. But our 100-gig strength is mostly on CDRs, and we haven't -- the penetration of CDRs in the less than 100-gig market hasn't been that great for us historically. So one of the points we've always stated about the 100-gig market is that most of those applications will need CDRs, whereas the lower-speed interfaces didn't necessarily need CDRs. So most of our success in data centers is coming from our CDR platform. So yes, we don't have a lot of exposure to the 40 gig. The strength in data centers is very good and is doing quite well.

Understood. And my last one and I'll jump back in line. So LoRa goes from handful of cities, 2 or 3, 3 to I think 12 now. Are you expecting a pretty seamless pickup in orders from -- associated with that expansion? Or would there possibly be kind of a pause as they try to ramp up for this large step-up?

Well, so first thing is they're going to roll out to 15 cities. So they started with 3, and they added an additional 12. So 15 cities is the stated number. And then the rollout, of course, you put the network in place. You put the gateways in place, and then that will drive some revenue, but most of the revenue will come from the sensors that are connected to the -- to those gateways. So it'll be kind of a follow-on effect. But I will say that Comcast is moving extremely quickly. They are doing -- adding a lot of specific verticals to these -- the markets wherever they see opportunity within the cities. So I think, as I had mentioned, obviously, our metrics, our number of gateways, number of sensors connected to those gateways, and as I've said, this year, we should do between $40 million and $50 million of LoRa revenue and then next year between $80 million and $100 million, and that's still the game plan I think.

Your next question comes from the line of Craig Ellis from B. Riley & Co.

I'll just stick with the topic of LoRa. Mohan, in your comments away from the Comcast agreement, you provided a very long list of activity that's occurring globally. Can you just summarize by helping us understand what that all means for the company's objective to take the business to $100 million in annual sales and where you see the business intercepting that revenue level?

Yes. So Craig, the reason why we talk about the different milestones in activity is because there are certainly bottlenecks in the industry to achieving our goals, right? And so we set up goals. In terms of the revenue goals for this year, we want to get to $40 million to $50 million of revenue and next year, $80 million to $100 million. And then I think if we get to that point, we'll see a path to potentially $0.5 billion to $1 billion of revenue for LoRa. So that's the reason why we set those milestones up, and it's really driven by how many gateways in each country and how many sensors connected to those gateways. And so we're monitoring, and I will articulate on every earnings call how many gateways and what our expectation is for the year and then what is driven by that. One of the key things about LoRa, of course, it's targeted at the IoT space and low-power WAN, and it's an emerging space. So some of these are very emerging new applications. There are some very exciting high-volume ones, and there are some which I would call more traditional kind of end-to-end connectivity or in spaces where the volumes may not be high but there are certainly cloud services and other values that the ecosystem can drive from that. So it's just kind of the overarching theme is that LoRa in general is now starting to become adopted across the globe as one of the key drivers of the LoRaWAN or the LPWAN market and the IoT space. And the types of companies that are involved, of course, with Comcast and Orange and SKT and ZTE and companies like that and Tata, you have significant influences in those regions driving LoRa. So it's a very exciting business for us.

Indeed. And if I could follow up, in the past, the company's talked about a licensing and royalty opportunity that relates to the business. As it relates to geolocation and any (inaudible) licensing activity with the business now at the midpoint of the year, what's your visibility into material licensing revenues for LoRa?

We have some good visibility, but I would say it's small today, Craig. I think, to me, the key thing is still, as we get more coverage and more gateways out there, then our service providers and our partners can start to really see some vertical applications that need geolocation and could use geolocation more effectively. And that's going to happen. I would say, as we talked about at our Analyst Day, we'll start to see that starting to ramp up, I think, next year. Early next year or maybe mid-next year, we'll start to comment a little bit on some of the applications and some of the revenues -- licensing revenues that we would expect from that type of application.

(Operator Instructions) Your next question comes from the line of Tristan Gerra from Robert W. Baird & Co.

Could you talk about the opportunities that you see for proximity sensing? And what's the environment that could push in various countries in the world for increased adoption for proximity sensing in smartphones?

Yes, Tristan. Proximity sensing is actually doing very, very well for us. We have good penetration of Korea smartphones, and it's starting to get penetration in the China smartphones. And it's essentially where -- wherever those phones are shipping into, where the regions of the world are starting to recognize that excessive RF emissions can be a health issue for humans. And so they are starting to put in regulations, and that drives the smartphone manufacturers to put in some sensing to sense the difference between the human body and materials like a table or something. So -- and so we're seeing that now start to pick up across the globe, and that really drives the smartphone manufacturers to look for solutions. And we think we have a very robust, very differentiated and very good solution. And so that business is really growing quite well for us. We're anticipating probably in the next -- it'll grow probably 40% to 50% this year over last year. Last year was a record. This year will be another record for sure. So yes, it's going very well.

Great. And then a quick follow-up on LoRa. I'm assuming that the average node per gateway is below 1,000, so that's well underutilized from a network standpoint. When do you -- how should we look at the WAN curve of nodes per gateway over the next year or so?

Well, so the capacity is about 5,000 nodes per gateway and it really depends on the network provider or the customers who are maybe the private end networks who are putting them in place and how quickly they want to roll out their sensors. It's difficult to say. I think some will push quickly to get as many sensors as they can connected. Remember, the key thing here is trying to get data, and you only get data if you connect the sensors. And that's the value of the network and the value of the end-to-end solution. So that's why we feel so confident that once the network is in place, the sensors will just -- it's just a matter of time whether it's 6 months, 12 months, 24 months, and then customers will just add more gateways as they need it. So I think that's the way to think about it that over the next 2, 3 years, we would expect all of those sensors to -- or the capacity, the full capacity of those 65,000 gateways to be in place.

Your next question comes from the line of Mitch Steves from RBC Capital Markets.

I have just one quick first one and then a follow-up. So first, kind of a long-term target. It sounds like the LoRa is growing faster than you guys expected and the mix is improving. Is there any sort of adjustment you guys need to make kind of [63 32] target you had back in your Analyst Day?

So Mitch, I'll just point, I think it is still too soon to start making any adjustments. Definitely, we are very pleased with how things are shaping out with regards to the targets that we reset about a year ago. But I think, at this point, it's still too early. Some point in the future, I'm sure we're going to have those conversations, but it's early at this time.

Got it. And then second, kind of returning back to Comcast. Is there any way to kind of quantify the acceleration and what you guys expected relative to what's happening now in terms of the rollout by city?

Well, the only thing I can say, Mitch, is that we had anticipated at the end of this year that they would make a decision and roll out to the first 12 additional cities, and they've already done it. So I would say they're 6 months ahead of what we had anticipated, and that is very encouraging from our standpoint because it really means that their trials proved to them and to us what we had anticipated but that they were going to be really good -- LoRa is going to be a right, good solution for them, and it was going to give them the type of benefit that they were anticipating. Otherwise, they wouldn't have rolled it out in the additional 12 cities. So that's the first encouraging thing. And then the second thing, which is really also encouraging, as I mentioned on my script, it isn't just about rolling out a network. It's about identifying the use cases and the ones that are potentially game changing and the ones that are very good ones from a business standpoint and can generate cloud services, et cetera, and Comcast is very aggressive on that front. So I think, on all fronts, it's -- I would say it's going much better than we had anticipated.

(Operator Instructions) There are no further questions at this time, and I turn the call back over to the presenters.

In closing, I am pleased with our strong performance for the first half of fiscal year 2018. We are benefiting from a number of secular demand drivers that are enabling us to deliver strong year-on-year growth. Our leadership positions in high-growth markets, including data centers, IoT and mobile systems should help enable the company to deliver record financial results this year. Our strong margins and spending discipline have enabled us to leverage our earnings growth well in excess of our revenue growth. With that, we appreciate your continued support of Semtech and look forward to updating you all next quarter. Thank you.

This concludes today's conference call. You may now disconnect.