Good morning. My name is Shelby, and I'll be your conference operator today. At this time, I would like to welcome everyone to the Teradyne Q2 2019 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star, then the number one on your telephone keypad. If you would like to withdraw your question, press the pound key. Thank you. Mr. Andy Blanchard, you may begin your conference.
Thank you, Shelby. Good morning, everyone, welcome to our discussion of Teradyne's most recent financial results. I'm joined this morning by our CEO, Mark Jagiela, and our CFO, Sanjay Mehta. Following our opening remarks, we'll provide details of our performance for 2019's second quarter, along with our outlook for the third quarter of 2019. The press release containing our second quarter results was issued last evening. We're providing slides on the investor page of the website that may be helpful to you in following the discussion. Replays of this call will be available via the same page after the call ends. The matters that we discuss today will include forward-looking statements that involve risk factors that could cause Teradyne's results to differ materially from management's current expectations.
We encourage you to review the safe harbor statement contained in the earnings release, as well as our most recent SEC filings. Additionally, those forward-looking statements are made as of today, and we take no obligation to update them as a result of developments occurring after this call. During today's call, we'll make reference to non-GAAP financial measures. We've posted additional information concerning these non-GAAP financial measures, including reconciliation to the most directly comparable GAAP financial measure where available, on the investor page of our website. Between now and our next earnings call, Teradyne will be participating in investor conferences hosted by KeyBanc, D.A. Davidson, Citi, and Bank of America. Let's get on with the rest of the agenda.
First, Mark will comment on our recent results and the market conditions as we enter the third quarter. Sanjay will then offer more details on our quarterly results, along with our guidance for the third quarter. We'll then answer your questions, and this call is scheduled for one hour. Mark?
Hello, everyone, and thanks for joining us this morning. In my prepared remarks, I'll highlight where we stand at the first half of the year, update you on current business conditions and our outlook for the second half in our test and industrial automation businesses. First, I'd like to welcome our new CFO, Sanjay Mehta, to the call. Sanjay brings a deep background in the semiconductor industry, along with direct expertise in ramping fast-moving global businesses. I am delighted to have him on our team. Sanjay will review the financial details of the quarter, we'll take your questions. We delivered financial results in Q2 that exceeded the high end of our sales and non-GAAP profit guidance by 3% and 2% respectively.
For the first six months of the year, compared with the same period last year, sales grew 4%, while non-GAAP EPS grew 15% on lower share count and higher gross margins. As was the case in Q1, the principal driver of outperformance in Q2 was strong demand for semiconductor testers for the emerging silicon used in the 5G infrastructure buildup. The 5G infrastructure investment represents the early innings of a multiyear buildup that will eventually work its way into mainstream handsets several years down the road. While trade tensions and sanctions have made it difficult to forecast and have created a volatile environment, we've yet to see any meaningful slowdown in demand for our semiconductor testers.
I will also note that the U.S. administration's actions in May to put Huawei and its affiliates on the Entity List has had no material impact on our sales. As other companies have announced, we performed an extensive review of our products sold to Huawei to determine whether or not they are subject to the Export Administration Regulations and the Entity List restrictions. Many of our products are not subject to the imposed restrictions. As a result, we have continued to supply these products to Huawei and its affiliates. For the products that are subject to the imposed restrictions, we are seeking licenses under the U.S. export regulations. The overall market picture in SemiTest has improved from our earlier view.
Entering the year, we estimated the SoC market size would be in the $2.3 billion-$2.7 billion range, down about $500 million from last year's market, based on forecasted weakness in automotive and industrial end markets, and an expected downtick in mobile device test demand. While auto and industrial markets have slowed as expected, the growth in tester demand for 5G-related infrastructure and demand for a wide range of semiconductors used in smartphone handsets has exceeded our earlier estimates. The 5G infrastructure demand that drove Q1's strong results accelerated in the second quarter and was complemented by an unexpectedly strong demand for smartphone-related chip test capacity. Our long-held view that increasing complexity is driving a meaningful increase in test demand is becoming more apparent with each new device generation.
All indications are that this demand will continue through the third quarter, and we are expecting this to result in an SoC test equipment market of $2.6 billion-$3 billion in 2019. On the topic of 5G, let me provide a bit of insight on how we're looking at the overall market opportunity and where we are in the progression. From a high-volume perspective, we see four phases of 5G rollout. First, an infrastructure buildup for frequencies below 6 GHz, followed by a ramp of handsets operating in those same frequencies. One or two years later, a similar infrastructure buildup for millimeter wave frequencies with a subsequent ramp for handsets that support those higher frequencies. In terms of volume, we are primarily in the first sub-6G infrastructure phase now, driven mostly by China infrastructure.
With the remaining phases playing out and building momentum over the next two -five years. While the big millimeter wave ramps are still in the future, there was a very active millimeter wave development work underway across the industry. We've been in the middle of that development, and lead customers in both SemiTest and at LitePoint. Recall, LitePoint was first to market with a millimeter wave-capable production tester, and just last week, we announced the MX44, our SemiTest millimeter wave product, which has been in the hands of early customers for several quarters. The MX44, an instrument for our UltraFLEX platform, delivers the RF performance, software ease of use, production worthiness, and economics necessary to support the development and early production of millimeter wave devices and modules.
In memory test, we continue to expect the market to be at the low end of our six to $700 million range. Our shipments in Q2 of $58 million were up over 20% from Q1 on the strength of flash final test and flash and DRAM wafer test growth. Despite the soft overall market, we continue to see strong demand for our NAND flash tester, as the push for higher speed interface testing continues to be strong. In industrial automation, Q2 sales grew 20% compared with Q2 of 2018. For the first half, sales grew 27%. The global malaise in industrial markets, especially those related to automobile production, were a stiff headwind. The automotive supply chain remains a large component of cobot sales.
Weakness in this sector has been difficult to overcome. Sanjay will provide more details on this sector in a moment. Given the lower year-to-date growth, we now expect full-year sales growth to be under the low end of our long-term range of 30%-40%. While below our goal this year, we remain confident of our long-term growth targets. We see this year's slower growth as a natural ebb and flow of the market and not a competitive issue as our results compare favorably to the broader industrial automation market, in which most companies are reporting sales declines compared with last year's Q2. We continue to work on ramping our lead generation activities, new market vertical diversification, and delivering initial availability of our TAM-expanding bin picking solution later this year.
While our IA growth moderated in the quarter, there were several significant milestones to note. Those include UR's largest account crossing the 1,000 robot install base threshold, continued progress on our large account program, and an OEM partnership with SACMI a leader in plastics manufacturing automation. At MiR, we now have several customer sites operating MiR fleets in excess of 25 mobile robots, and our install base of robots at Chinese hospitals has reached over 80 units and continues to grow. Before leaving IA, I'd like to comment on our recent investment in RealWear, an innovative private company that uses the power of advanced wearable technology, in this case, a head-mounted augmented reality device that makes the workplace safer and more productive.
This investment aligns with our strategy of bringing the power of advanced automation to companies of all sizes to improve the productivity of their employees and the quality of their products and services. Through the investment, Teradyne will gain insights into a wide range of applications and enabling technologies with potential use across our entire business. Back to the total company level. When you look at the full year, we expect the second half revenue and EPS to be slightly above the first half, with test a bit stronger and IA bit weaker than we expected three months ago. With that, I'll turn things over to Sanjay.
Thank you, Mark. This morning I'll make some comments on the first half of 2019, go through several highlights related to the business units, offer some observations about Teradyne after my first three months on the job, and then move to our second quarter results and third quarter outlook. We're pleased with our first half performance. As Mark noted, at the midpoint of our fiscal year, our first half sales totaled $1.058 billion, up 4% from the first half of 2018. A non-GAAP EPS of $1.20, up 15% from the first half of 2018. Gross margin improved one point to 58% in the first half of 2019, primarily driven by favorable product mix in SemiTest. The fundamentals of our SemiTest business remain strong, as noted.
With increasing test complexity and acceleration of 5G infrastructure investments in the marketplace driving demand, industrial automation continued to grow, albeit slower than expectations, but still outpacing the market, facing several headwinds. Turning to the business units. SemiTest had a strong second quarter with sales of $375 million. The key drivers of growth were, one, continued pull-in of testers supporting 5G infrastructure, and two, 20% quarter-on-quarter growth in our memory business, driven by DRAM and NAND test shipments. We expect 5G and memory demand drivers to continue in Q3. Regarding 5G, we see accelerated infrastructure spending for test equipment continuing in the second half of 2019. While we expect growing 5G handset-related test spending next year, we are forecasting a larger spending ramp in 2021.
Our LitePoint business grew 42% quarter-over-quarter to $41 million, driven by the system test requirements for new wireless standards and early 5G handset buying. We expect this level of business to continue into Q3. In system tests, revenues grew 25% quarter-over-quarter to $73 million, driven primarily by storage testers for multi-terabyte capacity and hard disk drives, and increased defense and aerospace shipments. Turning to Industrial Automation business. Our Q2 revenues were $75 million, which grew approximately 13% quarter-over-quarter and 20% year-over-year. As Mark noted, this is despite the glowing automotive investments, which is Universal Robots' single largest market. UR's 10% year-over-year growth to $63 million was less than our forecast. Geographically, growth in China and Asia Pacific remains relatively strong, but has been offset by slower growth in Europe and North America.
Even in the recent slowdown, we believe fundamental demand drivers of the cobot market, specifically the scarcity of labor, enhanced quality, financial returns, and unique ergonomic benefits in manufacturing will continue to drive long-term growth. We continue to believe we will go from an installed base of tens of thousands to hundreds of thousands of cobots in the midterm. In the short-term, we are seeing several industry headwinds working against our growth. Several quick market reference points will help calibrate the situation. First, the Robotics Industry Association, or RIA, is an association focused on the entire North American robot market, reported a year-over-year decline of all robot units sold of 29% in March this quarter, a trend we believe extended through the June quarter.
T his is principally due to the slowdown in the automobile manufacturing sector, which is over 50% of the North American robotics market. Global PMIs, a proxy for industrial growth, have moderated with 30 of 42 regional PMI measures around the world now below 50, thus indicating purchasing managers view that market conditions are contracting rather than growing. Obviously, we're not immune from the industry-wide conditions. For UR, we believe in the market regardless of the short-term headwinds, and will continue to invest to leverage our strengths in the product portfolio. Ecosystem and channel to invest in the product portfolio to extend our competitive differentiation. Lastly, we continue to train and bring on new partners and support lead customers in key vertical segments.
Recall, our strategy is to go to market with our channel partners, but to also develop relationships with the key leaders in large market verticals. These direct relationships enable us to understand the end market requirements for products, accelerate deployment time, and enable new solutions with third parties. Demand will continue to be fulfilled through our channel partners. In our view, the recovery coming out of market troughs provide an opportunity to extend our competitive lead and hence our continued investments. Later this year, investments will yield an industrial bin picking product with ease of use, flexibility, and economics that our customers have come to expect from Universal Robots. We expect new functionality like bin picking will increase the addressable market for UR cobots by approximately 50% plus.
This investment and others will position us for above-market growth when industrial investments re-accelerate. Shifting to MiR, which is obviously much earlier in its life. We continue to see healthy sales growth with second quarter sales of approximately $11 million, up 81% from last year's Q2 on a pro forma basis, due in part to the introduction of the MiR500 and MiR1000 pallet moving autonomous robots. While I've been on the job for just a few months, I'd like to offer some general observations. Teradyne has a well-positioned core test portfolio with secular market growth in the low single digits. The business model is flexible, with variable compensation tied to profitability, which ensures that all employees' objectives are aligned with the company's objectives.
Manufacturing is mostly outsourced for the core test business, which reduces CapEx and brings flexibility in our sometimes volatile market. We can focus on product road mapping without the burden of managing factory assets, which could hinder optimal roadmap planning and tie up significant capital, lowering long-term strategic flexibility. In May, I visited several of our contract manufacturing partners in China and Malaysia. I was impressed with the quality of our joint processes and mechanisms to ensure flexibility of production levels. We operate in markets that are volatile, and the flexible manufacturing operation delivers short customer lead times while modulating spending relative to demand. The other relevant point about manufacturing in China and Malaysia is that it enables us to minimize the impact of the current trade environment.
I've observed in industrial automation, we have a very different situation that we're defining and developing new markets. As these markets are in their infancy, we are focusing to stay well ahead and fortify our competitive position. Capturing these opportunities is one thing. Being able to scale to support these opportunities is another. Scaling manufacturing, a global operation, global distribution, and application ecosystem and so on, are areas where Teradyne has delivered synergies with our recent acquisition. It's also clear to me that our approach to integrating new businesses into the company is effective. To capture the market and realize the opportunities, we enable IA businesses to have decentralized operating decision-making. This allows our acquired companies' industry expert to continue to drive at an entrepreneurial pace with minimal bureaucracy.
Synergies are enabled through Teradyne's expertise in key support functions to drive efficient scale like supply line management, operations, legal support, global HR, and design for quality that all combine to accelerate growth. Which has sometimes hindered some smaller companies to truly scale and capture their market. There's a true collaborative management approach between business and corporate leaders that is focused on supporting the needs of these fast-growing businesses. A key difference between the core SemiTest business and Industrial Automation is manufacturing. As stated, our core test business has outsourced manufacturing for many reasons, including cost, flexibility, scale, and diversity of geographic location, something that is proving very important in these times. Our Industrial Automation portfolio is vertically integrated with manufacturing in-house.
Manufacturing in-house is a key differentiator in these new markets that we are trying to grow and enable fast time to market and quality solutions. Turning to capital allocation now. We'll stay disciplined and maintain the financial strength to return capital to shareholders, along with making acquisitions where it makes financial sense. Over the past three calendar years, we've averaged $420 million of annual free cash flow, which supports a balanced capital return approach with share buybacks, dividends, and acquisitions. We target maintaining approximately $1 billion on the balance sheet, earmarking $500 million to ensure we can ride out an economic downturn and continue to invest in our roadmap. This is paramount to our long-term success because when the market turns to growth from such a downturn, we will be well-positioned with a competitive roadmap to capture the opportunity.
In addition, we earmark $500 million for potential acquisitions to support our M&A pipeline. We also have $460 million in long-term debt in the form of a convertible bond due in 2023. We annually review our capital allocation approach with our board and will communicate any changes to you in the January call. Turning to the balance sheet. Our cash and marketable security stands at $994 million, about flat to the end of the first quarter. We returned $106 million of capital in the second quarter, principally through $91 million in share repurchases and $15 million of dividends. Our share repurchases since 2015 total $1.7 billion at an average buyback price of $30.44. Recall, we plan to buy back $500 million of stock this year and return $60 million in dividends. Turning to the second quarter.
Company revenue of $564 million came in slightly above the high end of our guidance for Q2, mainly driven by the acceleration of tester shipments for 5G infrastructure. We had two customers greater than 10% of sales in the quarter. Non-GAAP gross margin was 58%, non-GAAP operating profit was 24%, and non-GAAP EPS was $0.66. You'll see our non-GAAP operating expenses were $190 million, up $11 million from the first quarter as planned, primarily due to increased distribution investments at UR and some R&D expenses in the SemiTest, along with higher variable compensation tied to higher profits. The detailed segment level sales for the second quarter, including the geographic breakdown for UR and MiR, are shown in the table in the presentation. We continue to scale up our operating expenditures for Industrial Automation businesses.
We've included a schedule showing this breakdown between Test and IA. Let me mention one GAAP item of note. In the quarter, we had discrete tax expense of approximately $15 million related to the finalization of our repatriation tax bill liability. Turning to our guidance for the third quarter. Revenue is expected to be between $540 million and $580 million. The non-GAAP EPS range is $0.64-$0.74 on 171 million diluted shares. Q3 guidance excludes the amortization of acquired intangibles, restructuring, and non-cash convertible debt interest. Third quarter gross margin should run approximately 58%-59%. Total OpEx should run from 33%-35%. The operating profit of our third quarter guidance is forecasted to be 24%-26%. Shifting to taxes. Our non-GAAP full year tax rate is expected to be 16%, which is consistent with our prior guidance.
In closing, we've seen above forecast performance in our SemiTest business, driven by 5G infrastructure shipments. Our Q3 growth in SemiTest is mainly driven by a continuation of 5G infrastructure spending. Industrial Automation growth is slowing due to economic headwinds, but we expect the growth to keep outpacing the market, and we will continue to invest to drive leadership in product, ecosystem, and channel for our IA portfolio to achieve our midterm objectives. With that, I'll turn the call back to Andy.
Thanks, Sanjay. Shelby, we'd now like to take some questions, and as a reminder, please limit yourself to one question and a follow-up.
If you would like to ask a question, you may do so by pressing star one on your telephone keypad. Once again, that's star one. Your first question comes from John Pitzer of Credit Suisse.
Yeah, good morning, guys. Thanks for letting me ask the question, and thanks for all the detail in the prepared comments. Mark, just talking about the Industrial Automation business, I'm just kind of curious, you say it's going to be below the low end of your long-term growth rate target for the full year. How much below the low end, and how should we think about kind of half-on-half growth in the Industrial Automation business in the context that overall business is going to be sort of flattish half-on-half?
Yeah. I think it's obviously a little difficult for us to project with any precision the second half. Lead times are pretty short in IA. I think first you should expect to see an uptick in the second half in IA, kind of similar to what you've seen in past years, first half versus second half. We're going to see growth over the first half and the second half, and it'll be proportional to what we've seen so far in past years. In terms of where we'll end up in the year, in that 30%-40% target that we set, I think we're going to be below that, as I mentioned. I'd probably give it a wide range, somewhere in that low mid 20s to up to that 30 number.
That's helpful. As my follow-up, just on the SoC test TAM, you're raising it by about $300 million. I'm assuming the vast majority of that is 5G infrastructure, just given where we are on the handset cycle there. I'm just kind of curious, as you think about the four phases of 5G that you talked about, what's your kind of view on what it does to the SoC test TAM over time?
Yeah, we've talked that through a bit in the past. We've said that when we're up and running full speed, meaning the majority of the 1.5 billion world handsets have millimeter wave capability embedded, we should see about a $300 million-$400 million increase to the SemiTest TAM. In my remarks, I sort of said that's the last phase of this thing, and we've always said that's sort of out 2021 plus when that happens. We're still in that early infrastructure for primarily sub-6 GHz deployments. Yes, there's here and there scattered deployments of millimeter wave infrastructure in the U.S., but nothing compared to sort of the much more major rollout going on in China for sub-6 GHz.
Thanks. Appreciate it.
Your next question comes from Brian Chin of Stifel.
Hi there. Good morning. Thanks for letting us ask a few questions. Sanjay, welcome to the company and the call. I hope you got a glass of water a fter your remarks. First, just curious about the third quarter breakdown relative to the revenue guide. Roughly speaking, how do you expect your key business segments to trend in 3Q relative to 2Q?
Yeah. It's Sanjay here. Specifically, I think the SemiTest business will continue along with the drivers of the 5G infrastructure. As Mark mentioned, you should see an increase in the industrial automation, similar to what you've seen in increases in the past. I think some of our other businesses, like LitePoint, the growth was really driven by the new wireless standards, the Wi-Fi 6 or 802.11ax, seven gigahertz would be examples, you should see continuance on that front. On the System Test group, which did grow 25% kind of quarter-over-quarter in Q2 to $73 million, you should see continued growth there. Really from our hard disk drive business, really driven by the data center, the enterprise disk drives going into data centers.
As well as the defense and aerospace business, which are typically large deployment programs with government agencies, and some large ones are moving into the deployment phase.
Okay, got it. Very helpful. Appreciate that. I guess for my second question, maybe kind of a two-parter. First, in terms of looking at UR versus MiR, I guess UR was decelerated a little bit in the quarter to +10% year-over-year growth. Kind of curious sort of what that monthly progression was in the business and/or by geography. Did you see more of a maybe a pronounced tail off towards sort of late in the quarter? I guess also MiR, a little surprising, kind of still earlier phases of adoption clearly for that business, some headwinds, but maybe just any commentary you're also see ing just in terms of sort of that business and sort of having to maybe revise down some of those contingent payouts in Q2.
Yeah. I think from a UR perspective, I think we're seeing, you're right, 10%. Really when I think about the markets of, in my prepared remarks I commented on Europe and North America, you're really seeing the impact of the slowdown tied to automotive. Many other of the robotics companies are actually showing negative year-on-year results. However, we did see strength in Asia Pacific. When I think of China, Korea, and Japan. We're looking forward, we actually expect that to continue to be relatively strong. We'll have to get back to you on the monthly profile. I don't have that handy.
Okay.
Just maybe a comment on MiR a little bit. MiR has had a very ambitious plan tied to its earn-out from day one. As we sit here in Q2 and look at where we think the full year will turn out, originally to sort of max out in this year, they would have had to have roughly doubled in sales. Where we sit now, given the first half results, is we're probably going to be in the certainly much greater than 50% range. The doubling potential, given some of the headwinds that Sanjay mentioned, is less likely.
Okay. Thank you.
Your next question comes from CJ Muse of Evercore.
Thank you for taking the question. First question on the SoC side. You had previously talked about second half tracking maybe $50 million lower than the first half. Is that still in the cards here? As you think about moving into 2020, and obviously, over the last few years, we've seen very elevated SoC spending, and that's in spite of the headwinds we saw from Apple two years ago, now this year, auto industrial, how can you kind of quantify rising complexity versus some of those headwinds, and how we should think about that translating into a market size into 2020?
Yeah. Let me give you a few guide points on that. I do think because of the strengthening SemiTest environment we're seeing, that first versus second half SemiTest revenues are probably going to be about flat. As I mentioned earlier, IA should be up. That gets to the comment that I made that the second half should be a bit stronger than the first. The complexity drive of the business. I think when you look at the two things that are happening this year, despite the fact that handset unit volumes are declining and have essentially been flat to declining for several years now, we still see a very robust uptake of demand for semiconductor testers. Obviously not unit driven, complexity driven.
The complexity drivers in the more recent years, this year and some others, have been around, a lot of that is related to the increase in the number of and the density of the cameras that are going into phones. Some of the higher-end phones coming out this year all have six cameras in them. That propagates throughout the phone in terms of complexity. The amount of NAND flash you need goes up, the speed with which the NAND flash has to operate goes up. Same thing with the DRAM. All of this is part of the complexity story for cell phones that drives our business, and we're yet to get into the high bandwidth 5G related silicon that's coming. That's sort of the thesis we've had.
I think the evidence so far is it's playing out pretty well.
That's helpful. I guess a follow-up on the IA side, as you contemplate the weakness that you're seeing today, at least relative weakness, can you kind of pinpoint where that's coming from, between auto, European exposure, China exposure, perhaps U.S. companies deciding to buy other components before potential tariffs put in place? Would love to hear your thoughts there.
Once again, I think Europe and North America are really tied to the slowdown in the auto industry. Just to talk a little bit, we still believe we're in a nascent market, and once again, we did grow overall the 20% in the quarter. In these n ascent markets, we continue to invest in different applications for things like bin picking that will drive a wave of adoption, or we're investing in new market verticals like hospitals. For example, with MiR, there's roughly 80 mobile robots deployed in hospitals. Then we're also investing in large accounts. These large accounts typically have a little bit longer design cycle and qualification process as they're evaluating different competitors. Again, we really see a big market going forward.
We believe we're helping to create and drive this, and what we're looking for are the market to re-accelerate its spending to then pull the products forward.
Thank you.
Your next question comes from Timothy Arcuri of UBS.
I had two. I guess the first one, Mark, is the comments around the 5G infrastructure. I think that's pretty consistent that ultimately you think it's going to be $300 million-$400 million incremental to the TAM.
Hey, Tim, we lost you there, Tim. Could you repeat the question?
Sure. Can you hear me now, Andy?
Yes.
Yeah, my question on 5G infra. It's $400 million increments. It's just very consistent with what you said in the TAM. I'm surprised that you're raising the TAM of SoC as much this soon. I know sub-6 GHz is kind of thinking through this plan and should we really be projecting this type of a TAM in next year?
All right. Tim, you were in and out there, but I'm going to try to paraphrase the question as you're wondering about our raising of the TAM this much this early and attributing it to 5G. Is that indicative of a longer-term step up in the market size?
Yeah, I think that's what Tim was asking, hopefully. Look, we're looking into Q3. We still see very strong demand, a lot of it related to the infrastructure. I think raising the TAM at this point has a good solid basis. I think the other part of your question was how much legs does that have into next year? I think it's too early for us to tell. There's a large build-out going on in China, and there's plans for that to continue over the next several years, in fact. It's not like it's going to be over at the end of this year. On the other hand, there's all kinds of other economic conditions and tariffs and things that c ould temper that.
I wouldn't go out on a limb yet talking about specifically next year, other than to suggest that the early innings of this, as I said in my comments, are generating a significant uptick in the SemiTest market, and that should allow us to build confidence in our mid-term earnings model that is actually relatively modest in market growth. We feel pretty good about what's happening this year as a positive proof against that model.
Awesome, guys. Thank you. Yeah, that was my question. Thank you. There's another question. I think, Mark, you also talked about better smartphone. I'm a little bit surprised about that as well. Is that your biggest customer? Can you give us a little more comments there? Thank you.
Yeah. I'll just say what I said before. I'm not going to comment on any specific customer, but the thing in smartphones, one of the key things has been the proliferation of more cameras, denser cameras, more pixels, more test time. Without unit growth, that complexity growth, and that extends into memory, is one key element of why we see a strong smartphone semiconductor test market this year.
Okay, guys, thanks so much.
Thank you.
Your next question comes from Toshiya Hari of Goldman Sachs.
Hey, good morning, guys. Thanks for taking the question. I've got two. On memory test, Mark, I think this was the first quarter in a long time, it may be the first quarter ever, your memory test revenue exceeded that of your nearest competitor in the quarter. I realize it's only a quarter, you're clearly making good progress on the market share front. Curious if the strength was driven more by your traditional NAND business or increasingly your DRAM final test business as well. I've got a follow-up.
Yeah. I didn't expect to see that in my lifetime, and I don't expect that will persist. Look, I think what we've been talking about consistently is still the case, that the growing sub-segment of memory test tends to be gathering around the high-speed variants of DRAM and flash. What we saw in the second quarter was a continuation of this sort of NAND flash final test business we saw in Q1. A lot of the adder is due to wafer test. It's the wafer test of both DRAM and flash final test. If you said quarter-on-quarter growth, what's the main component of that? It's the wafer test piece.
Got it. As my follow-up, on LitePoint, I think a couple of years ago when business was really slow, at one point you guys were losing money in the business. I think you had some cost-cutting initiatives. Since then, revenue has improved. Curious, where do margins sit at LitePoint today? I guess more importantly, going forward, when do you expect 5G to become a meaningful driver for that business specifically?
Hi, it's Sanjay. I'll take that one. LitePoint is profitable, and you should think about it a little bit above, currently running a little bit above our company average. Indeed includes shipments to support 5G and will continue to grow.
Thank you.
Okay. Next question, please.
Your next question comes from Mehdi Hosseini of SIG.
Yes, sir. Thanks for taking my question. I have a follow-up. Mark, you talked about incremental TAM increase of $300 million, $300-$400 for 5G sub-6 GHz phones. This year, your SoC TAM went up by the same amount, $300 million. Should I assume that the base band station, the networking TAM is about $300 million, and when the phones are out, that adds another increment of $300-$400?
No, I wouldn't think of it that way. The $300 million-$400 million that we talk about for the 5G business incorporates phones, base stations, and the infrastructure associated with 5G.
Sure. Okay.
It's all one thing.
Okay. This year is all about base station, and you increased the TAM by $300 million. Does that mean that when the phones are out, the content or the demand drivers changes from base station to smartphone?
I think, first of all, the three to $400 million TAM increase this year isn't exclusively base station. There's also smartphone silicon that I've talked about that's higher than our expectation, related back to a lot of these image sensor trends. It's a combination of phones and base stations, what we're seeing this year. As you project out, over time, the base station piece of this should have good legs for several years to come. Everything else being equal, there should be continued growth for several years to come. That will taper off, maybe it's around 2021, 2022, and the handset piece will take over the main driver.
Let me try one more time as the follow-up to my first question. Of the $300 million-$400 million TAM associated with sub-6 GHz 5G, how would you characterize the base station opportunities or the mix, and how does it compare to a smartphone?
It depends on what year you want to talk about. If you want to talk about the next three years, it's not going to get all the way up to that $300 million-$400 million rate. It'll be below that, and it'll be primarily driven by base stations. After that sort of two to five years out, it will get up to the $300 million-$400 million adder, and at that point, it will be mostly base stations. Let's say in the next two couple of years, it's, let's say, $200 million-$300 million. After that, it's $300 million-$400 million, and there's a shift from infrastructure to phones.
Sure. Thank you. My second question, could you also help us quantify opportunity associated with 5G impacting LitePoint? I assume that that's more of an old phone. Is there any figure you can give me as it impacts your board level test or LitePoint?
Yeah. We've talked about that business, too. We said that adds about $100 million to the market. It's mostly phones, but there's also a lot of, with Wi-Fi 6 and coming 7 GHz band Wi-Fi, there's a lot of access points and infrastructure there that will also grow. That's probably 20% of LitePoint's business, but it's meaningful.
Okay. Thank you.
Your next question comes from Krish Sankar of Cowen and Co.
Yeah. Hi. Thanks for taking my question. I have two of them. First one, Mark, is there a way you can quantify how much of your SemiTest and/or SemiTest plus LitePoint is coming from 5G today? What % of revenues? Then I had a follow-up.
Yeah, we really don't break that out. Maybe we'll look at that to see if we can't do that going forward. At the moment, we really haven't consolidated that.
Got you. No worries. As a follow-up, on the LitePoint side, is there a real opportunity for LitePoint in the frequency range 2, FR2 or, I guess, millimeter wave, or do you think there's opportunity in FR1 also for LitePoint? Along the same path, I think I asked this question last time, too. It seems like there are two key players in SemiTest, but there are four players in wireless test. Do you think if that industry on the wireless LitePoint side consolidated, there's better opportunity for everyone involved? Thank you.
Yeah. FR1 is less of an opportunity for all of us than FR2, but there is still growth as FR1 rolls out. The competitive dynamics in that business, we've talked about it before, it's a crowded market. The thing that LitePoint specializes in is on the production test optimization. Their products really are not designed for R&D purposes. Therefore, I think in the production test market, although there are four or five competitors for development tests, there's truly fewer for production tests. We may be talking three instead of five. Yeah, it's still a little bit crowded, but I think for production tests, a little bit less crowded than you might expect.
Got you. Thanks, Mark.
Your next question comes from Atif Malik of Citigroup.
Hi, thank you for taking my questions. Good job on another beat and raise quarter on 5G strength. Mark, we hear in Taiwan that you've entered probe interposer market. Can you just talk about the strategic rationale and how big that opportunity is? I have a follow-up.
Yeah, I'm not exactly sure what you're referring to. Certainly interposers are part of the sandwich that makes the tester dock to the wafer. We haven't made any announcements around any specific products there, and we're not ready to talk about any of that.
Sure. On IA side, are you seeing any retaliatory action by China in preferring local cobots that makes you look at this business different strategically?
Yeah. Hi, it's Andre. Actually what we're seeing in China is actually a return to growth. What we're finding is that customers are taking a look at the competition. I think as it was mentioned on the last call, what they're finding is that hardware is still a differentiation, because to basically run the cobot 24 by 7, kind of three shifts, we can sustain the performance from a hardware perspective, not to mention the software, the ecosystem, and the other benefits. Even from a hardware perspective, we continue to outperform the competition. Now, that's not to say that the competition isn't coming. We are seeing competitors come and be around specifically in China. If anything, we actually have a little bit more of an enhanced view as the year has unfolded.
Okay. We'll take the next question, please.
Your next question comes from Richard Eastman of Baird.
Yes. Good morning. The first question is just around the industrial automation business. A couple things there. One is, are we seeing any incremental traction on kind of these enterprise agreements or direct sales? We had booked a couple of those, I think, in the first quarter that you spoke to in the lighting industry. I'm curious if we have any more examples there. Also within IA, as we coast through the back half of the year, is the EBITDA target? Do we slow investments, maybe? Is the EBITDA target for the full year in IA still expected to be 16% or better?
Yeah. It's Sanjay. I'll speak to the EBITDA targets and kind of what's going on. First thing is, again, I'll reiterate, we believe in the long-term of this market, and we're going to continue to invest. Really, you should think about it as a trade-off, where we're going to continue to invest if we believe in the mid and long term, we're going to reap benefits from a revenue perspective in those investments. Our first priority is to make sure we're developing competitive differentiation, either in product channel or the ecosystem to drive revenue. We'll forego a little bit of operating profit through those investments to obtain that, really from an investment standpoint. I think in 2018, roughly our operating profit was 16%. What you should expect in the short term is that that is going to be ±.
In the long term, we do expect that our investments will be leveraged and will grow towards the company average.
Just on the large account discussion. Yes, in the first quarter, we added a couple of large accounts that I was describing that were rolling out cobots in sort of this 20-ish to 30-ish units per month rate. Those customers continued to perform and continue down that path. We had this other large account that I mentioned in my remarks that's now up over 1,000 robots. That's occurred, by the way, over about a three-year period. They're obviously rolling out at a much higher rate than that. We didn't have anything of that magnitude additionally in Q2 to talk about in ter ms of somebody else who's in that 20 to 30-a-month rate. There are those in the pipeline, hopefully. Our plan would be to talk about those, if we get permission, in coming quarters.
Okay. Mark, just as a follow-up question. On the SemiTest side of the business, there were a couple spots flagged there in auto and industrial. Could you just maybe speak to what % of SemiTest are targeted at those industries and maybe what tester product line we could look to see that exposure?
Let's talk about automotive first. Traditionally, automotive has sort of been this 400 million-ish test market for us. Mainly, it breaks into microcontrollers and say, engine control systems, and then power analog, the actual actuators. It's split between our J750 tester line and our Eagle tester line. Over time, as the electrification of vehicles has been moving forward and complexity is increasing, more and more of those devices are finding their way onto our UltraFLEX SoC platform. There's a migration occurring. As we've talked about automotive before, we saw three very strong years of automotive demand that was sort of unprecedented from 2016 through 2018. This year, the demand is maybe down 40% from what it had been running at for those three-year average. It's been a pretty significant pause. That's normal.
I would've expected it earlier than it occurred, and it typically doesn't last much more than a year and a half at most. I think that's fine. In the industrial space, similar products, product lines. It's the J750 moving to the UltraFLEX and the Eagle test platform. The trends there are very similar, maybe not down as hard as auto. Perhaps it's only down 30% or so compared to where it's been running the past couple of years. In a similar vein, we expect that to last for about a year or so, and then there's a recovery after that.
As a percentage of SemiTest, 15% or?
Again, if automotive's $400 million out of an SoC, let's say, market that's nominally $2.6 billion or $2.7 billion, you can do the math, the linear one is probably more nominally $300 million, $350 million t he industrial.
Yep. Okay, great. Thank you.
Your next question comes from Sidney Ho of Deutsche Bank.
Great. Thanks for taking my question. I want to go back to the 5G question. You talk about the impact of the Huawei ban has not been significant for you, or you don't expect impact in the future. If you look at the purchases from them, specifically for the 5G infrastructure, are you seeing them buying in line with their build plans? How can you tell if that customer is not pre-buying for future quarters?
I'm not going to comment specifically about any one customer. I will say that whether or not any customer is kind of buying ahead of demand or not is something we're always trying to triangulate on. If we look at what we saw in the second quarter in terms of buying, there's no evidence that there was any buying ahead of demand in the second quarter. We look at that constantly. Like you said, we triangulate shipments of end products out with what we know we're supplying in and see if that all adds up. All I can say is that through the second quarter, we don't see a disconnect.
Okay, that's helpful. Maybe another end market question other than auto industrial. You refer to better than expected growth in 5G infrastructure in your press release, but you also mentioned strength in networking. Are we talking about the same thing or is it a different type of networking? If you can give some color, it'd be great. Thanks.
Yeah. It's probably a subtle distinction, but it's essentially being driven by the same thing. You have in a 5G rollout, you have the radio access network, which has antenna modules, down converters, then you have modems, and then you have a backhaul to network processing. The networking we talk about is sort of the connection of the backhaul through the network processing related to 5G infrastructure.
Okay, great. Thank you very much.
Operator, we have time for just one more question, please.
Your final question is from David Duley of Steelhead Securities.
Thanks for taking my question. Just a clarification on the size of the SoC market. I guess you increased the size of the TAM this year. Could you just in reference, how big was the SoC market in 2018 and this year you expect it to be, I guess, $2.7 billion, is that what you said?
Yeah. Last year, the SoC market had a phenomenal peak year of about $3 billion in size. At the current midpoint, we're talking about $2.8, $2.7, something like that. It's still down. It's down maybe that 10%-ish range.
If the market's down 10%, what will we expect Teradyne's SoC test business to do this year?
Teradyne's SoC test business this year, I think go back to what I said, we're not specifically guiding the fu ll year there, but we expect our second half to be roughly equivalent to our first half. You can divine it from that.
Okay. Final question. You mentioned a couple different times about test time intensity. I imagine you're referring to your APU customer or any sort of complex chip like that. Could you give us an idea of generation over generation, what sort of increase in intensity you are seeing?
Yeah. It depends, really. It's hard to have a rule of thumb there, because first of all, generation to generation, the devices themselves obviously get more complex and have more transistors. Everything else being equal, test time would go up generation to generation. There's a good correlation between transistor count and test time. However, now and then certain things happen to optimize the test methodology. That could be some architectural thing in the tester that allows more efficient testing. You might find a generation where the device got more complex, but test time didn't go up. It could be a test technique or a quality issue that improved on the customer side. In general, with highly complex digital silicon, we kind of see that 10%-20% natural migration in test time offset by some of these other one-time events.
Final thing, did you mention who your 10% customers were? I can obviously guess who one is, I was curious kind of about who the other one was.
No, we didn't. We specifically won't call that out in the quarter. However, at the end of the year in the 10-K, we will provide disclosure and just remind you that customer buying patterns are what I would call a little lumpy. Having a 10% customer in the quarter doesn't surprise me. We will provide that disclosure at the end of the year, should they be greater than 10% for the entire year.
Can you help us out as far as end market goes or any sort of color?
No, I don't think we're going to get into that, but if obviously this situation persists, we'll be talking about it as we get into next year.
Thank you.
Okay. That about wraps it up, folks. Thank you for joining us. As a reminder, if you have follow-up questions, please reach out to me directly, and we appreciate you joining. Take care.
This concludes today's conference call. You may now disconnect.