Cohu, Inc. (COHU) Earnings Call Transcript & Summary

Good morning or good afternoon, and welcome. My name is Jeff Jones, and I'm the Chief Financial Officer for Cohu. We're glad you could join us today, and we look forward to walking you through the Cohu growth strategy and new financial model. Before I go further, I'd like to call your attention to our safe harbor statement and remind everyone that this presentation is being recorded and will be available for future viewing in the Investor Relations section of our website at cohu.com. Today's presentation will last approximately 60 minutes, followed by about 30 minutes of Q&A. Our CEO, Luis Müller, will start by sharing an overview of our markets and strategies for differentiation and growth. We're fortunate to have 2 guest speakers today that will discuss their company's market-leading technologies and how Cohu adds value to their test process. Our guest speakers are: Michael Campbell, Senior Vice President of Product and Test Engineering at Qualcomm; and Ganesh Moorthy, President and Chief Operating Officer of Microchip Technology. Ganesh has recently been appointed the next CEO of Microchip, starting in March. Congratulations, Ganesh. Then our general managers will provide more details about our strategy for growth and how we differentiate ourselves in some focused markets and applications. I will conclude the prepared presentations with an overview of our financial results and new midterm business model, and then we'll have about 30 minutes for Q&A. Now I'll turn it over to our President and CEO, Luis Müller.

Thanks, Jeff, and welcome, everyone. Yesterday, we issued a press release in which we increased our Q4 guidance and Q1 outlook due to strong order momentum, particularly in mobility and automotive. Jeff will get into more specifics about this later. I'll begin today's presentation reviewing our new midterm strategy and target model for growing revenue and profitability. We are also fortunate to be able to share the voice of our customers directly with you to illustrate how our solutions fit into their semiconductor manufacturing process and deliver value to their business. Cohu is a fast-growing company in semiconductor test and inspection, an industry estimated at $6.5 billion. At the midpoint of our revised Q4 guidance, Cohu will deliver revenue of approximately $631 million this year, capping a 3-year compound annual growth rate of 21%, which far exceeds the industry average. We have an active installed base of over 28,000 systems across many blue chip customers where we have built strong partnerships. Cohu's revenue composition is unique, with about 45% coming from a more stable base of recurring business. Our midterm 3- to 5-year plan is to grow revenue to $940 million and operating income to 23%. The main pillars of the Cohu growth strategy consists of extending our leadership in the high-growth RF test segment, serving a $400 million addressable market, solidifying our strong handler position with cutting-edge thermal and vision technologies, improving customers' yield in a $700 million market with secular growth applications in automotive and mobility; increase in the contactor attachment rate to our semiconductor equipment in a $500 million market by leveraging our significant installed base; and enhancing test cell differentiation and building a competitive moat. Finally, we plan to provide a suite of software and data analytics to optimize test cell productivity and increased recurring business. We're focusing on select end markets that are making semiconductors more ubiquitous to our everyday lives. We believe that the industry is at the beginning of a multiyear growth period. 5G is a disruptive new technology that will connect everyone and everything, delivering orders of magnitude faster data exchange. 5G penetration in smartphones is only 12% today and projected to more than double next year and continue growing for several years as millimeter wave proliferates. RF content in smartphones is forecasted to grow by 50% in the next 3 years. This will open the gates to new applications, not only in mobility, but also in industrial and medical, automotive and computing, with an estimated 40% increase in global data creation, transmission and storage over the same time period. In automotive, where Cohu has a leading position with thermal handlers, digitalization with ADAS processors, sensors, telematics and the electrification of the drivetrain will transform the way we move around. Market researchers forecast an 18% annual growth rate for automotive semiconductors over the next few years. We will focus on these secular and multiyear trends to achieve above industry revenue growth, delivering differentiation in a $2.5 billion serviceable market. Cohu's one-stop shop solution enables our customers to achieve faster production yields and optimize equipment productivity with best-in-class RF test instrumentation. The most responsive active thermal control solution in the market, utilizing artificial intelligence to maximize inspection yield and data analytics to optimize productivity. We aim to capture $35 billion or more of incremental test cell revenue next year in addition to growing in each of our served markets. We're well positioned to capitalize on these multiyear mega-trends targeting 14% annual revenue growth, 2x the industry rate, and 23% operating income, accelerating share gains in test and inspection with disciplined R&D spending that delivers 30% or higher return on invested capital. At the target model, the business is estimated to generate $160 million annual free cash flow. I hope you find the presentations that follow informative as they bring to light our plans and product differentiation. Next, I want to introduce Michael Campbell from Qualcomm.

Good morning, good afternoon, and good evening. This is Michael Campbell. I'm Senior Vice President of Product and Test Engineering at Qualcomm. We're here today to talk to you a little bit about Qualcomm and how Qualcomm works with Cohu. The world everywhere is becoming a 5G world. Everything needs to be connected and interconnected intelligently. That drives bandwidth and the need for more speed and more data at all point places and points. 5G allows us to build the bandwidth needed to accommodate your parking meter, your robo dog, your car and everything else that's being interconnected these days. Qualcomm is the leading wireless technology innovator that has been working for 35 years to build the patents of basic technologies and the strength to be able to continuously compete. We started off with CDMA moving analog to digital, move into the smartphones and desktop. And today, with 5G, we're going to connect everything virtually everywhere from your car to your -- to the cloud, to AI, to your Internet of everything. Qualcomm's R&D engine is a $64 billion cumulative R&D engine with over 140,000 patents and 300 license agreements. Our world-class technology portfolio includes wireless connectivities, processors, multimedia components in the RF front end. Today, Cohu is our partner with their tester for the RF front end, the advanced thermal handling capabilities for our processors and multimedia testers. Our business segments at Qualcomm are really the semiconductor business segment, which, in 2019, was about $14.6 billion and our licensing group in 2019, which was about $4.6 billion. Qualcomm's growth vectors are really tied to 5G, scaling that 5G and 4G business into the adjacent business structures and growing new opportunities. VR glass is the factory of the future, automation, AI are all things that we're working on for the future that will change and transform how the world connects, computes and communicates. 5G wireless becomes a superior technology that will disrupt everything. It's as fast and as reliable as cable. With that very fast infrastructure, we can go from sub-6, which is basically a straight extension of the current LTE networks to millimeter wave. All that will give us the bandwidth and the capabilities to connect everything, driving up to a $13.2 trillion economic impact of 5G-enabled goods and services in -- by 2035. We're not just in another handset upgrade cycle. 10x decrease in end-to-end latent activity, 100x increase in network efficiency are just two of the spectrum ideas that drive the -- that 5G drives when it's fully enabled from LTE extensions into the millimeter wave. 5G is accelerating globally. 85 operators commercially deployed. 310 additional operators are investing in 5G. There will be 2.8 billion connections in 5G by 2025. This is a huge market we're about to immerse in and very robust across the whole gamut of -- from factories to your phone, of applications and uses. The digital transformation that we're going to be driving with 5G goes from health care to manufacturing from agriculture to energy. All these bandwidth, all these items require cloud communications or AI communications to be able to make them better, more efficient and more effective. 5G allows the cloud to the device to through the AI connectivity that drives improvement in whether your traffic signals or your retail signals or your energy signals. When we break through, not if we break through the technology barriers, we're going to leap drive and leap the ecosystem forward with us. Today, a great example is streaming multimedia. Today, you can stream multimedia, but it's got glitches quite often, snags and lags. On your cellphone with 5G, there won't be those things. We will be streaming Netflix, Prime Video, Apple Video, Spotify and Hulu seamlessly. The question is, what will the consumer do with it, and how will the consumer change the bulk delivery vehicle and the execution vehicle for the 5G? The truly immersive VR experience with your smartphone will allow you, Steve, to sit at home and enjoy the video from your house in the comfort of your own home, but you will be there because you'll be able to see that the key cameras in the marketplace that are streaming the video, you'll be able to see the stage, you'll be able to hear the sound and you'll be able to feel the intense stadium all around you while you're on your 5G phone. 5G phone is also going to enable you sports experiences that have been never seen before, real-time sports with real-time stats, play-by-play action being screenshot to your cellphone and enabling your cellphone to be basically your own artificial intelligent announcer for the sports. The car is part of the 5G network. The 5G network will enable cars to be more intelligent. It will tell the cars where consumers are, where there's bikes, where the other cars are and what traffic patterns are. That will enable smoother traffic flow and will enable an easier to work, from work and automatic cars one step closer to reality. 5G is also making the age of virtually everything smarter a reality. Whether you're making your home smarter by having your home connected to your cellphone or connected to your office desk or to whatever, your AI assistant, the world will become a smarter place. Cohu, our partner, has helped us with MSMs, which are our basebands, our digital basebands by using their active thermal-controlled handlers to keep the temperature accurate. Why is that important? The temperature coefficient of these very complex, very dense, high-speed devices is critical to keep on an accurate basis. The Cohu's active thermal control allows them to or allows us to both improve yield and improve the consistency of the product we're marketing for PCs and other more complicated devices. That temperature is more accurately measured with Cohu's equipment and allows us to drive yields up and test times down, thus making the product more robust in the marketplace. Same thing for the RFPAs products. Cohu's RF subsystem and the cost point in the marketplace allows very accurate and robust testing of RF components. And allows us to bring that marketplace, especially for 5G, to the market much easier. So those two things come together. Qualcomm using Cohu's products and leveraging Cohu's technology and Cohu leveraging us by figuring out where our road maps are going, so they can build better technology looking forward. In this age, the real age of 5G made -- concrete made practical by Qualcomm, everything starts to interconnect. We're going to be driving enabled consumers. We're going to transform industries, enrich lives, and we're going to invent breakthrough technologies were around this bandwidth and the infrastructure that we have that will change the world. So with that, I'd like to say this is the age of 5G, made real by Qualcomm and being helped by our key business partners. Thank you very much, and I hope this talk was of use.

Hello, everyone. My name is Ian Lawee. I'm the Vice President and General Manager of Cohu Semiconductor Test Group. I joined Cohu in 2019 after 25 years in the industry, initially at Teradyne and most recently in Analog Devices, where I led multiple semiconductor product and end market segments. Cohu's recent results highlight how increasing test intensity is fueling our growth. In 2020, we saw a strong ramp of first-generation 5G and Wi-Fi 6 RF semiconductors. As Michael shared with you, the world is becoming increasingly interconnected. Many new applications and devices need higher data rates and lower communication latency. This need is achieved with more frequencies in wire bandwidth per IC, increasing test times up to 25% and the average system selling price by an average of 30%. These trends allow us to sell new systems and upgrades to our current installed base. We are projecting to grow revenue 18% annually in mobility, leveraging our leading position in RF front end IC test and targeting new applications with our universal platform. Multiple factors are contributing to both near-term and future growth outside of mobility, particularly in the automotive and industrial segments, where we are forecasting 13% revenue growth annually. Automotive and industrial RF are benefiting from the introduction of new standards like ultra-wideband and with Industry 4.0 increasing number of distributed sensors. Furthermore, proliferation of advanced packaging is leading to more complex mixed-signal and power management devices. This trend is increasing the adoption of structural test, which is an early step in the test flow, typically at wafer growth. For these insertions, our air cool test platforms of our cost effective solution, including for complex devices, such as application and AI processors. Through the second half of 2020, we executed on the fastest ramp yet of a new instrument, our RedDragon RF instrument suite. This third-generation instrument suite covers all the RF technology inflection points I just discussed, and is designed for the wave of new standards and technologies being adopted now and in the near future. As Michael described, our customers depend on us to provide accurate and robust systems and to make it easier to deliver their complex 5G technology. That means accelerating the time to ramp at high yields and delivering solutions that address our customers' entire device portfolio. The RedDragon instrument suite covers Wi-Fi 6 and sub-6 gigahertz and millimeter wave 5G devices while maintaining backwards compatibility to 4G and older WiFi standards. By designing the RedDragon into our well-established platforms, customers can upgrade existing systems in the field at a fraction of the cost of a new test cell, strengthening our advantage over competitors and better positioning Cohu to sell incremental capacity to test 5G devices. Cohu has the scale to enable high-volume production ramps and simultaneously support our customers worldwide engineering and production facilities. We have seen a number of design wins this year, where Tier 1 customers selected RedDragon over their existing rack and stack in-house systems, recognizing the need for our next-generation platform that better aligns to RF front end IC test requirements as the segment grows from $15 billion to $25 billion over the next 5 years. The RedDragon RF suite is available on our packs and Diamondx platforms. Cohu's universal test platform provides the broadest price performance range, spanning from a true 0 footprint, 5-slot configure tester to a system that delivers hundreds of devices tested in parallel. Our platform strategy has a strong advantage for RF devices, including RF transceivers for IoT. It is also well aligned to power management, display drivers and microcontroller device testing. The Cohu universal test platform is also suited for the growing structural test applications I discussed earlier. These are device categories that cut across many applications and benefit from tailwinds in the automotive, industrial, consumer and computing market segments. Cohu is expanding our value by integrating differentiating technology in the test cell. An example is millimeter wave where tester to contact or calibration is key to performance. Our customers historically had to manage this challenge themselves through design and cross vendor coordination. As a technical challenge grows with millimeter wave, so does the difficulty of homegrown solutions. A Cohu test cell reduces the risk and effort, thus delivering improved yields and faster time to market. 5G and Wi-Fi 6 are driving significant growth this year and in the future as the technology adoption spreads and frequency requirements expand. We have planned our RedDragon instrument suite to deliver that capability modularly, so customers can add functionality as needed. We are also seeing the early adoption of new instrumentation in automotive and industrial connectivity, radar and a growing satellite Internet applications. We have leading RF test solutions today for these emerging markets. And we are investing to meet the needs of what we believe will be a growing $400 million RF test market in the next 2 to 3 years. So to summarize. We are very focused on the fast-growing device segments that are best served by our platform strategy, delivering differentiated solutions to a $1.1 billion addressable semiconductor test market. We are growing with 5G and IoT RF markets that are transitioning from rack and stack in-house systems to production testers. We have visibility to at least $20 million of incremental revenue in the first half of next year from share gains in this segment. Given our strong global RF expertise, a proven established platform and the ability to upgrade a large installed base, we are confident this will continue to be a strong component of our semi test business. We are also extending to applications in automotive, industrial and computing markets. As you'll hear from Ganesh from Microchip, our platform is well positioned at leading customers, and we are investing to add instrumentation to extend our advantage for testing other devices such as OLED and microLED display drivers, power management devices for electric vehicles and data centers. The semi test revenue growth target of 17% is driven by our differentiated test instrumentation, universal platform and Cohu's unique position to solve challenging customer integration problems and deliver a test cell solution. Thank you. Next, you will hear from our second keynote speaker, Ganesh Moorthy from Microchip.

Good morning or good afternoon, everyone. My name is Ganesh Moorthy, I'm the President and Chief Operating Officer of Microchip Technology. Microchip is a customer of Cohu for many years, and it is my distinct pleasure to speak to the Cohu investor conference today. By way of introduction, Microchip is a leading total system solutions provider providing high-performance microcontrollers, microprocessors and digital signal controllers, analog of many forms, a mixed-signal, power, interface, security solutions, clock and timing, which are universal in all systems, wired and wireless connectivity, FPGA solutions, nonvolatile memory and Flash IP solutions. Our fiscal years, which end in March. In March 2020, the fiscal year ended with $5.3 billion of revenue. We're headquartered in Phoenix, Arizona in a city called Chandler, and we have about 18,000 employees worldwide. Our enduring vision for the last 25 plus years has been to be the very best embedded control solutions company ever. That has meant different things at different stages of the company. Today that is built on 3 pillars: how do we enable systems to be smarter and smarter. How do we add more value to the smart systems when they're connected. And how do we ensure that a connected system is secure. As we think of our road map of total system solutions, we look at this from the point of view of our clients, what is it that they're trying to build. Obviously, you can see here in the center in the dark is the intelligence inside the system. But that alone is not enough. You need the signal chain that brings the information to the intelligence. You need power. You need clocks, which are universal in all systems. You need to have all the communication requirements, the security requirements. We build all of these building blocks, either as discrete or integrated onto our microcontrollers themselves. We have had a remarkable track record of growth and profitability starting from 1993 when we went public and over the last 27 plus years. During this whole time frame, we've got 120 consecutive quarters of profitability, including some of the time when we were private. And that is an unmatched track record for anybody in the semiconductor industry. They're good times, through bad times, we've been rock-solid in profitability, allowing us to be strategic in our investments, in our support for customers and our innovation, in our capacity, et cetera. And you can see how this has tracked over time with different product lines, microcontrollers and processors in blue, which are about 55% of the company. Analog and mixed-signal, which are about 28%, 29% of the company in orange. FPGA, which is a more recent addition, about 7% to 7.5%. And then everything else, about 10% or so. Our growth has been fueled by a relentless gain in market share in the microcontroller business, which is the business we've been in the longest. This shows you that remarkable progress from 1991 when we entered the market as a small underfunded company in a field of giants and the consistent shares that we have gained over this time frame, reaching #3 in the 2017-2018 time frame, and continuing to close the gap with #2 and 1. Today, we're only 16% away from #1 and growing faster than both #1 and #2. And it is our full expectation that we will cross them somewhere in the next 3 to 5 years. While the world knows us well for our microcontrollers, we are less well-known for analog, but we are gaining substantial share in analog, too. Here's the data for 2019, showing that for the first time, we're in the top 10 at $1.5 billion of revenue, the only analog company in top 10 that recorded growth in 2019. This becoming the second engine of growth for Microchip. As we look ahead, we have 6 end markets that we play in, shown on the left-hand side of this slide. But equally, importantly, we have 6 megatrends that we see providing durable growth for the next 5 to 10 years. Those 6 megatrends are the advent of 5G and the rollout of that infrastructure over the next 5 to 10 years. IoT, specifically as it pertains to the industrial IoT, the insatiable growth of data and what it does to driving data center growth. Electric vehicles becoming a higher and higher percentage of the total vehicles in the world. Artificial intelligence and machine learning from the cloud to the edge to the end nodes themselves. And last but not least, automated driving and autonomous driving. These are 6 outstanding growth trends that we intend to capitalize on in the coming years. Looking at our business by end market. The largest end market we play in is the industrial end market, followed by data center and computing, followed by automotive and then smaller segments in communication, consumer appliances, in particular, aerospace and defense. So industrial and automotive, which is a focus for this conference, are really two very large market segments for Microchip. So what do we mean by the industrial market? It's a very fragmented front end applications that are pictorial represented here for some of the applications, but it really represents hundreds of other applications, things that go into the factory infrastructure, things that go into the cities and other infrastructure and really represent a great, long term, sticky business. Takes a long time for us to develop and a long time over which that revenue will last. In automotive, too, we are designed into a wide range of applications all the way from body control to human machine interface to USB to networking to touch interfaces to audio systems. In an average car, we have somewhere between 30 and 40 Microchip products designed into it. A high-end car, probably in the neighborhood of about 70 to 80 components of Microchip that are designed into it. Shifting now to manufacturing. Unlike most semiconductor companies, we own and operate a significant amount of manufacturing capability. About 20 different factories around the world, 6 of them, which are very high-volume factories, producing, on average, about half to maybe slightly more than half of the output for the company. It is in the area of manufacturing that Cohu begins to intersect Microchip and our strategy. Now one of our guiding values is that suppliers are our partners, and we started our partnership with Cohu 15 years ago. We learned quickly that Cohu takes the time to listen and understand our needs and to partner with us to develop creative solutions that end up exceeding our expectations. Let me share with you one of many examples. Almost every microcontroller we sell has high-performance analog circuits. Cohu listened to our need for an analog resource that enables testing a very large number of units at the same time that resulted in a new tester board. I will show you the benefits of this collaboration on the next slide. Cohu understands that changes to capital equipment strategy often require a long-term view, design and development of products and hardware, quarters to years to complete. And Cohu realizes that more engineers that are trained when using their equipment and the more designs they enable on their equipment through their applications teams result in future purchases. They have made investments in application support in both good times and bad times to help both of us. We also appreciate that Cohu recognizes the benefits of the embedded control solutions that Microchip offers and incorporate into their own designs. The result of our partnership with Cohu is that they are one of several suppliers that have enabled Microchip to improve the efficiency at test by enabling highly paralleled test solutions as much as 80%, 90% reduction in the footprint that is required and the labor that is required. And while the results will vary by product family, Cohu's test solutions have allowed us to implement these on many product families and make significant manufacturing improvements in the process. In conclusion, our market leadership requires clarity of vision, continuous innovation, consistency of purpose, relentless execution and, of course, great partners to enable our business. We are proud of our partnership with Cohu and expect this to continue for years to come. Thank you for the opportunity to speak to you today.

Hello, everyone. My name is Chris Bohrson. I am Senior Vice President and General Manager of Cohu's Test and Inspection Handler business. As you have seen in the opening video, Cohu develops and supplies a broad portfolio of handlers with unique thermal, vision, contacting and mechanical stimulus for testing semiconductor devices and sensors. Our systems are used both in test and scan as well as stand-alone vision inspection applications in the semiconductor industry. I would like to start by discussing the evolution of packaged technology and how it influences handling and inspection. The two segments where we expect the biggest impact in growth are automotive and mobility, including inspection. As you just heard from Ganesh Moorthy, in automotive, the two trends, turbocharging growth or electrification in ADAS or autonomous driving. Included in the growth of ADAS or higher powered microprocessors, microcontrollers and sensors. These trends are driving changes in how devices are tested. For ADAS devices, there are 2 significant impacts. First, while these devices have traditionally been tested at extreme temperatures ranging from minus 55 Celsius to plus 175, in order to support ever-increasing quality standards, these devices now must be tested with tighter temperature variation and up to 32 in parallel. The compounded challenge, ADAS microprocessors also operated higher power and safety during test. So in order to maximize yield and minimize test time, the handler most dynamically controlled device temperature while rapidly removing heat at very high temperatures. Also electrical vehicles required to operate on higher voltage. And testing and higher volume production without working requires a specialized testing environment. Addressing these challenges will enable Cohu to grow test handler revenue in the automotive segment at 21% compound annual growth rate in a $300 million market segment. About half of this growth will be driven by returning to normalized levels from where we expect 2020 to close and about half driven by the changes I have just outlined. Moving to mobility and inspection. To support device miniaturization and smartphones, the industry is increasing the use of wafer-level chip scale packages, or WLCSPs. With tighter quality standards, it is critical that our customers can identify defects in these advanced packages, including subsurface microcracks. The challenge is to differentiate real defects from cosmetic problems and reduce overall rejection. Also sensors are getting more complex, particularly the latest generation of microphones that require precision testing and high-volume production. Overall, mobility and inspection are major segments for Cohu, and we are targeting to grow revenue in test and inspection handling at 7% compound annual growth rate in this $400 million market segment. I would like to now focus on how Cohu is addressing these challenges and translating them to higher yields for our customers. Cohu has a long history and is the industry leader in active thermal control, or ATC, for testing high-power devices. ATC enables our customers to dynamically control the temperature of the device during high-volume testing in order to make sure they can minimize test time and maximize yield. We cut our teeth in this space with the world's leaders in computing and in the past 6 years, have extended that to mobility. As I pointed out earlier, the challenge is now to provide this in the automotive space for higher power microprocessor. Cohu's solution for this is T-Core. T-Core technology is widely deployed on a range of systems around the world in computing, mobility and now automotive. In head-to-head benchmarks, T-Core continues to provide best-in-class active thermal control for high-volume applications as well as our customers' most challenging devices. For advanced inspection of WLCSPs, the challenge is to find real defects and meet the stringent quality requirements of our customers. I emphasize real defects is without the right capability, it is very easy to over reject parts due to cosmetic problems. These do not affect performance but create a nuisance as traditional inspection system is mistaken for real defects, thereby reducing yield. Cohu's NVcore Vision technology provides best-in-class inspection in the standard and high-resolution visual light spectrum. Also, we have extended our unique capability further to include true infrared inspection and deep learning algorithms. True infrared inspection provides the ability to see micro cracks on and below the surface of the device. Cohu has also developed deep learning algorithms to distinguish scratches from actual defects. Cracks potentially impact performance while scratches may not. The challenge then is to distinguish between the two in high-volume production. Cohu's NVcore Vision technology has unique capability proven in high-speed production to find real defects and minimize over rejection. The benefit of this to customers is that by ensuring quality but reducing over rejected, this protects our customers' quality standards while increasing yields. Beyond T-Core and NVcore, Cohu provides other capabilities that also deliver better testing and higher yields. First, our system for testing many sensors enables handling of delicate devices at very high speeds. Also, the Cohu architecture isolates testing of sensors from environmental influences, including vibration and noise. A good example is testing precision microphones where we need to minimize the noise floor for accurate measurement. Second, our ability to provide more or all elements of a test cell is a significant Cohu differentiator. Two examples of this include that we can combine T-Core with Cohu's PROTRACE pro trace contacting technology to extend our thermal performance in test cell applications. Further, our solution for testing high-voltage devices is a handler configured with an inert gas chamber combined with Cohu contactors to create an arc-free, high-voltage IC test in volume production. Okay to summarize for the test and inspection handler business at Cohu, our served available market is $900 million. We plan to grow revenue at 12% compound annual growth rate by delivering leading edge capabilities to segments that we serve today, including computing, industrial and consumer while focusing on higher growth, automotive, mobility and inspection opportunities. Cohu's growth in automotive will be based on our unique ability to address our customers' technical challenges with emphasis on the high growth segments. This includes more semiconductor content in automotive, including engine management, safety and infotainment. Growth in automotive will also be turbocharged by new IC content supporting ADAS and electrification. ADAS is driving the need for active thermal control for high-power microprocessors that require dynamic heat removal during extreme temperature testing. Cohu is uniquely positioned to take advantage of this with its production proven T-Core technology, providing active thermal control. We will extend this advantage by providing solutions combining handlers configured with-t core technology and Cohu's PROTRACE contactors. Also, electrification is driving growth in the semiconductor content in vehicles and creating demand for high-voltage testing. Deploying handlers that utilize specialized test chambers combined with Cohu contactors will provide our customers with high-quality, high volume, arc-free testing for this application. In mobility and inspection, growth is based on Cohu's best-in-class and NVcore Vision inspection technology that delivers more accurate inspection, including infrared. We further enhanced defect recognition with artificial intelligence algorithms to maximize inspection yield. Finally, we have developed a unique architecture to isolate sensor testing from the environment, enabling precision measurements required for advanced microphones and our applications. As a result, we deliver value to our customers with lower test costs and higher yields by providing differentiated solutions for fast-growing applications in automotive, mobility and inspection. So that concludes my update. Thank you for your time. And next, you'll hear from Devin Sheridan, General Manager of Cohu's Interface Solutions Group.

Hello. My name is Devin Sheridan. As Chris indicated, I'm the General Manager of Cohu's Interface Solutions Group. I joined Cohu in August of this year, and I'm very excited to be working with a dynamic and highly innovative team. Previous to Cohu, I spent over 27 years in the semiconductor business in various roles from process engineering to global operations management in test and new product introduction. More recently, I was Director of Operations for Asia new product introduction and test innovation at NXP. Now moving to the presentation. The Interface Solutions Group develops innovative technologies to solve complex test challenges. Differently than what Ian and Chris previously described, Cohu's Interface Solutions group sells consumable products to semiconductor manufacturers and as such, create a steady recurring revenue stream that is not based on test capital expenditures but on the customer need to keep production running day in and day out. The mobility segment represents a high-growth opportunity for us, driven by expanding 5G adoption. Within this market, initial growth has been primarily in sub-6 gigahertz applications. However, we see a growing number of devices and applications expanding to higher frequencies, known as millimeter wave. The millimeter wave device applications represent a significant increase in complexity, supporting a growing number of RF bands. Additionally, at millimeter wave frequencies, managing signal integrity and efficiency through the entire hardware stack from the device to the tester instrumentation becomes very challenging. The complexity around test causes the cost per channel in the test instrumentation to grow exponentially. To address these challenges, customers are requiring innovative interface designs and a higher level of integration to efficiently balance technical complexities with test economics. In the automotive and industrial markets, semiconductor companies will continue to benefit from growing silicon content in vehicles. The highest growth coming from vehicle electrification of an estimated 3-year 27% CAGR. And ADAS with an estimated 3-year 29% CAGR. The automotive market is characterized by stringent quality requirements that demand robust and reliable solutions. Growing silicon content in vehicles will demand 0 defects where time to yield is highly valued. Test must innovate around electrical and thermal control as well as integrate within yield and data analytics to enable 0 defect performance. For electrification, this requires greater precision for improved battery and power management. With increased performance requirements, customers expect higher test parallelism to control costs. As Chris mentioned earlier, ADAS processors require active thermal control across high parallel test to minimize test time and maximize yield. Here an integration between handler and interface hardware is needed to accurately manage temperature across the full test array. Devices in the computing and networking market, which target cloud, AI and edge processing use state-of-the-art process technology in advanced packaging. As computing demands increase, both pin count and interface speed requirements stress the limit of contacting technology. We believe this represents a significant opportunity to disrupt interface market with innovative solutions enable testing of high bandwidth, large-scale account devices while driving improved cost of test. Cohu's interface products provide innovative solutions that maximize performance, particularly when combined with our handlers and testers in key market segments. For 5G, we developed contactor technologies that control impedance and inductants for applications up to 50 gigahertz. Our solution extend standard contactor components to higher frequencies with the use of engineered materials and designs, providing a high level of efficiency for our customers. For frequencies above 50 gigahertz, our xWave platform provides a direct signal path from the device to test instrumentation, critical for maintaining high signal fidelity. In high-performance digital computing, networking and ADAS applications, our innovative coax contactor combines exceptional impedance control with full signal isolation that is critical in today's high bandwidth protocols. Cohu's solutions are scalable to high pin count packages. For analog devices used in the industrial and automotive markets, we have a full array of precision measurement contactors that scale the high power, enabling test at peak voltages in high current that are used in electric vehicles in today's industrial applications. What we are most excited about is our ability to innovate beyond the standard contactor, who has developed a class of intelligent interface solutions under the PROTRACE brand that enables greater yield and operating efficiency for our customers. For applications with high thermal demand, we embed sensors into the contactor to enable fast calibration and closed-loop temperature control. For RF, we integrated millimeter wave switching capability to expand parallelism and enable calibration to the device under test. The lab combines improved cost structure with greater test accuracy, optimizing cost and efficiency for millimeter wave devices. Cohu is in the unique position to extend innovation across the entire test cell. This is a highly valued capability that reduces our customers' time to yield. This is especially critical for technically challenging applications and for customers with limited integration resources. Here, Cohu has attached the RedDragon RF-enabled tester with the PROTRACE intelligent contactors to extend millimeter wave frequencies to high parallelism with greater test accuracy. With our smart contactors, we can tightly integrate PROTRACE and our T-Core system to reduce mature variation across the test array and reduce calibration time. Cohu's test cell solutions improve test quality and accelerate the time to the volume manufacturing of customers' new semiconductor devices. Essentially, the Cohu test cell allows for the entire integration reducing time to market and delivering greater operational efficiencies faster for both semiconductor manufacturers and their test subcontractors. Innovative high-performance whole test cell solutions will provide differentiating value to our customers and enable Cohu to capture a higher market share. Our strategy enables our rate of attachment to double over the next 3 to 5 years. I am extremely excited about the opportunities in front of us and believe we have the right team, the right products and the right strategies to win. Our strength is in solving our customers' most complex as challenges. In applications that require high link and transmission speeds, we deliver conventional probe technology that is familiar, manageable and cost-effective for our customers. In mobility, this means we provide standard test economies to millimeter wave applications. And for computer networking, we bring exceptional signal performance to high bandwidth protocols. For automotive, we bring a long history of performance and quality to devices that must scale on computational capacity with tight thermal control and ADAS and supported precision power management required for electrification. Cohu's ability to innovate across the entire test cell, addressing both cost and performance is unique and differentiating strength that sets us apart in the industry. We are confident that our strategy will enable us to win and grow. We set our target to grow at 20% annually in this $500 million addressable market. Thank you. And now I'd like to pass the presentation over to Jeff.

Hello, again, I hope you found the business presentations helpful to understanding our customers' test and inspection challenges and the innovative technology we bring to address them and ultimately improve production yield. As Luis mentioned, yesterday, we issued a press release increasing our Q4 2020 financial guidance, driven by share gain in RF test and accelerating automotive demand. We now forecast Q4 revenue to be $195 million to $200 million, which at the midpoint is 31% higher than our Q3 2020 revenue. Gross margin to be between 45% and 46%. Operating expenses to be approximately $53 million and adjusted EBITDA to be approximately 20%. Please note that all these figures and my comments to follow are non-GAAP figures. Please see the presentation appendix for explanatory notes and reconciliations. Moving to the balance sheet. We have reduced our Term Loan B debt by nearly $21 million so far in Q4, executing on our priority to deleverage the company. Looking ahead, we now project Q1 2021 revenue to be approximately 5% higher than our Q4 updated revenue guidance. Now I'll walk through our recent performance and new financial model targets, highlighting the results of our strategic actions and operational execution. The estimated 2020 numbers reflect actuals through Q3 and plus the midpoint of our updated Q4 guidance. Our strategy has resulted in a revenue growth rate over the last 3 years of 21% and gross margin expansion of 300 basis points. We attribute this growth to organic development and acquisitions, resulting in share gain and high-growth RF test with scalable precision instrumentation, strong handler market leadership delivering differentiated thermal control technology to improve test yield, increasing contactor attachment rate by leveraging our handler market position and best-in-class technologies, delivering test cell solutions to accelerate customers' time to yield and nearly half of our revenue coming from a steady base of consumable and other recurring business necessary to support the utilization of more than 28,000 systems in our customers' test facilities. As a result of increasing demand for mobility and automotive test, we estimate Cohu's second half 2020 revenue will increase by approximately 23% compared to the first half of the year. On the cost side, prior to the end of 2019, we completed the actions necessary to reduce annual costs by $40 million. Operating costs are tightly managed, and the model has strong operating leverage as reflected in the expanded profitability estimated for the second half of 2020. Due mainly to improved operating leverage, we estimate second half of the year profitability will approximately triple the profitability in the first half of 2020. Today, we are introducing a new midterm target financial model that delivers higher profitability at all revenue points. Our new target model achieves $3.60 of annual EPS on revenue of $940 million versus $3 of EPS in our previous model. This represents a 20% increase in profitability at target revenue. On a quarterly basis, the new model reflects an increase in EPS of $0.10, growing to $0.15 at target revenue of $235 million. Gross margin expands with higher revenue from greater leverage of fixed costs in our manufacturing facilities and favorable product mix as our higher-margin semi test and contactor businesses grow at a faster pace than our handler and PCB test businesses. Higher profitability comes from further reductions in operating expenses. OpEx has limited variable cost and will continue to increase at a rate of about $1 million for every $10 million increase in sales. The combined effect is an improvement in operating leverage with about 45% of of incremental sales falling through to operating income. We have added free cash flow targets at each level of quarterly revenue. CapEx requirements are relatively modest and are modeled at about $20 million per year. We had a strong cash position at the end of Q3 at $171 million. 95% of our debt is the Term Loan B, which partially funded the Xcerra acquisition. The Term Loan B is covenant light, has favorable minimum payment requirements and matures in 5 years. Our priorities for cash generation and capital allocation are investments in technology and product development to support organic growth, consistent and meaningful debt repayment. We're targeting annual repayments of approximately $60 million, subject to business conditions and cash projections and acquisitions to expand our served markets and possible technology tuck-ins. Starting with our estimated revenue for 2020, the path to 940 million includes leveraging our market leadership and best-in-class technologies to improve test yield, as you heard in detail from our business unit general managers. Automotive test was low throughout most of 2020. However, we remain well positioned with our customers and have experienced a steep order recovery in Q4 that sets the stage for an excellent start to 2021 automotive revenue. The automotive test market is poised for multiyear growth based on projections for the proliferation of semiconductors required for electric vehicles and driver assist. In mobility, our FIC test demand accelerated in the second half of 2020 and and this market is projected to continue growing over the next few years as the number of 5G-enabled phones is projected to double in 2021 compared to 2020 and double, again, in 2022 compared to 2021. And as you heard from our general managers, opportunities for cross-selling Cohu products and providing the complete test cell is growing and creates competitive differentiation and barriers. Now why Cohu? Our strategy is to grow revenue profitably and faster than the overall market rate. We can achieve a growth rate 2x the overall market by focusing on high-growth end markets and increasing our customers' test and inspection yield with technologies and products that help solve increasingly complex challenges. We believe current market growth drivers, specifically for automotive and mobility, will exist for multiple years with increasing adoption rates of 5G phones, electric vehicles and automated driver assist. We have taken significant steps since the Xcerra acquisition to eliminate redundancies, streamline manufacturing operations and optimize our overall cost structure. As a result, our operating leverage continues to increase. The estimated fall through of incremental revenue to operating income is 45%, and we've increased EPS by 20% on target revenue of $940 million. Increasing profitability strengthens free cash flow that we intend to use for product development supporting organic growth, consistent and meaningful debt repayment and potential acquisitions to further expand our served market and/or technology portfolio. In summary, Cohu is well positioned to capitalize on growing market opportunities and deliver financial results in line with our midterm target model. This concludes our presentation, and now we'll move to the Q&A session.

[Operator Instructions] Our first question comes from Brian Chin with Stifel.

Thanks very much for hosting the event and for all the significant detail and disclosures. Maybe first question here. Just kind of zeroing into the model a little bit before I move to products. On a quarterly basis, it looks like Cohu is approaching or at approaching $200 million in sales, which annualizes out to roughly $800 million. Your 3-year plus annual sales target is $940 million, which is about 15%, 20% higher than this. So I guess my question is, could this 3-year target be either too conservative in terms of expected sales synergies, market growth or timing? Or is there some degree of cyclicality that you might be accounting for over the horizon?

Brian, it's Jeff. Yes. I mean, as you know, it's extremely difficult to forecast out even a couple of quarters. And so going over the next few years is virtually impossible. We know there is seasonality in the business. We know that in times of ramps, customers have -- had occasional positives while they digest equipment and reassess the market. So we feel pretty good about where we are. We've got a lot of momentum coming into 2021, and aiming for that $940 million target.

Okay. Got it. Maybe shifting quickly to the bottom line. One of the key updates here seems to be the tighter OpEx control at the higher revenue levels. And the better earnings fall through. I guess, if I run the 12% of sales R&D sort of target through the business, that implies, again, very minimal creep up in the SG&A expense dollars. And I think you said $1 million for every $10 million. If I think about it, that might be. But was that SG&A or was that total OpEx?

Total OpEx. Yes, total OpEx.

Yes. Total OpEx. So a really minimal creep up in SG&A expense. Maybe, can you just dig in on sort of how you again plan to achieve this sort of efficiency over the years? And then maybe also for Luis, in terms of the growth in the R&D dollars, how much of that is roughly to improve and strengthen within existing served markets versus to expand into new served markets?

Yes. So let me just give you a couple of points of reference, Brian. So we're targeting total R&D, which includes -- in this case, includes application engineering, which is heavier for our semi-test or tester business. So we're targeting total R&D to reflect about 14% of sales. Now that's on average. So it's going to be a little bit higher for some business units, lower for others. But we do indeed feel that that's adequate to develop products necessary for organic growth. Overall, the relationship is that OpEx grows by about $1 million for every $10 million in sales, and that's roughly split between the R&D and the SG&A.

And just to continue on to your second part of that question, Brian. The majority of that R&D is assigned to existing markets that we serve. Now as you heard or hear from the business units, in certain of those markets, we have a smaller share, and we're aspiring to grow faster. And so it's already a market segment that we supply to but there's a big opportunity ahead for us to increase our penetration and share in those segments. So that's where the R&D is going to, whether it is inspection or RF test or power management, IC, automotive, ADAS, EV, they're all segments that we currently serve.

Okay. That's helpful. Maybe my last question is sort of a three-parter on contactors. As a benchmark for Cohu's 20% revenue growth target for this segment, what do you expect the underlying growth rate for the contactor served addressable market to be over the comparable horizon? The second part of that is maybe the -- what is your current revenue profile for contactors now? I think you sort of distributed that across 3 buckets: mobility, auto industrial and computing. I would imagine there's a bit of a lean towards auto industrial, but if -- be curious what that profile looks like today. And then maybe to sustain your 20% growth, what attach rate is likely required to achieve this?

Yes. Okay, Brian. So I would say the current revenue profile, we give a -- on one of the slides there, as you noticed, a pie chart of the estimated breakdown for revenue for 2020. So currently at a rate of about $100 million for 2020. Maybe I'd turn it over to Devin to address the other 2 questions in terms of -- and of that revenue profile, we don't necessarily break that down into markets. Maybe Devin could give a little color on that. But in terms of your growth rate of the SAM, and then your attach rate to hit the target 20% growth rate, I'll turn that over to Devin.

Yes, Brian, thanks for the question. Yes. So I'd say that the auto market and mobile market are probably the largest ones for us, right, and the largest opportunities for us going forward. And they -- that's where a lot of our innovation is targeted. And that is going to drive -- help us drive to that 20%. Coming up behind that, I think, computing, which is not necessarily as large, but it's a large market. We also have some key innovations there that gives me confidence that can help support our revenue targets for the coming years.

And then, Devin, as related to the attachment rate, to sustain the 20% annual growth, we're looking at moving that up to about 50% to 60%. Is that correct?

Yes, that's correct. And so the key areas for us in attachment, right? So we have a strong track record in automotive that drives that. And so as that drives up, that's going to increase our overall attach rate. But really, the innovation that the team has, working with Ian and Chris and their organizations. And innovation and mobility, I think, really helps us there. And that's going to be a driver that helps build up our attach rate. And then like I said, even in the compute area, as we have a strong position in handlers. And so as we innovate there and bring in some of our products, that will help with the attach rates.

Got it. And just to clarify, in case I missed it. Again, the target bogey for the contactor business for Cohu is about 20% growth over the 3 years. Do you think the SAM there is growing 10%, 15% to sort of triangulate?

Devin, what's your estimate on that?

Yes. No, that would be correct. It's right around 10% is what we see the SAM growth there. And so we would see expansion beyond that with our 20%.

Our next question comes from Krish Sankar with Cowen & Company.

I had a couple of them. First one, either Jeff or Luis, on your long term or your midterm revenue target of $940 million in revenues, how does the composition of the revenue look like between 5G, auto, recurring, et cetera?

So Krish, this is Jeff. So let me give this to you. Let me look at it maybe a slightly different way, and we can get to your answer. So looking at the -- more or less the handlers that are -- have a heavier focus to automotive and industrial, it's going to be about 48% of that $940 million revenue. If we look at the semi test, the testers, which are -- be more focused on mobility, that would be about 30% of the business. And then contactors, which would be probably 60-40 between automotive and mobility, would be closer to 20% of the business.

Got it. Got it. That's very helpful. And then just to follow-up on the point, you kind of also emphasized how you're seeing a massive boost in orders on the auto side. And clearly, auto should be driving incremental revenue growth in calendar '21. I'm just kind of curious, I think you guys mentioned about a $400 million SAM for the auto semi test market. How would that grow as semicons in auto increases? Do you still think it's going to be a $400 million? Or do you think it could move higher down the road?

So why don't I turn that over to Chris, who talked about the handler and inspection business.

Yes. So the -- if I understood your question correctly, this is Chris Bohrson. The -- when we think about automotive, there are really 2 factors driving the growth, right? One is, obviously, growth in the vehicle demand. And we've estimated that to be about -- in 2021, to be about, I think, 13%. Then there's growth in content. And the content is really driven by several factors. Obviously, there's a growth in the all infotainment and safety systems and solution control and whatnot. But as I mentioned in the presentation, it's really being turbocharged by 2 factors, right? One is growth in electric vehicles. The other one is in automated driving. So we see the content growth next year of round numbers about 11%. And so that's really driving a major chunk of growth in the semi industry of up to 23% for next year. And I think Luis mentioned about an 18% compound annual growth rate in the automotive semi business going forward. And that's what's really driving our growth of the 21% that I mentioned as -- in that $300 million market segment for automotive.

And Krish, I'm just going to add that, that's very focused on the handler side, right? There's also a portion of the contactor business attached to automotive. And as Ian mentioned, also the semiconductor test business in automotive. Obviously, we're strongest there on the handler side that Chris just detailed, but the other pieces as well.

All right. That's super helpful. And then just a final question, if I could just squeeze in either for Luis or Ian or Chris. I think you guys mentioned how auto demand is high in, at least in Q4, the current quarter has gotten back to 2018 levels. I'm just kind of curious from the way when you look at it, what is the correlation between auto units in a given year to semi test opportunity for you folks? Is it a one-to-one correlation? Or is it more -- there's a lag effect? How should we think about that?

Chris, do you want to talk about it? And maybe Luis can add some color, too?

Yes. Krish, Chris mentioned during the talk there that we have 2 components going on, right? We have the IC content penetration in automotive, which is now actually running at hundreds of dollars depending on the automotive, whether it's an electric vehicle could be -- even going up to $500, $600 per car. And then there's obviously the unit growth in vehicles. The correlation is really a combination of those 2 factors, right? And we're looking ahead here. We have both looking pretty strong going into 2021, as Chris mentioned, I think the -- they're both running in the teens right now for 2021 and compounded sort of that kind of 20 -- I think it was a 23% growth for semiconductors in automotive for next year and 18% compounded over the next 3 years. So it's not a one-to-one to unit vehicles, but it's more of a one-to-one to a combination of unit vehicles plus content growth.

Next question comes from Tom Diffely with D.A. Davidson.

First, Jeff, I just want to confirm all the numbers that we saw today as far as growth rates go are based off of the 2020 full year numbers, that's growth from this year?

Yes, that's right, Tom.

Okay. So I'm curious then, if you look at a cycle-over-cycle, maybe peak-to-peak or trough-to-trough level, how different would those growth rates be versus the ones that you gave us today -- starting today going over the next 3 years?

I would say that we're in a fairly unique period here with the tailwinds that we have. I would say that if we look back over the recent cycles...

Yes, I took a look at the -- over the last 10 years. And it looks similar to the coming off of 2009, for example. I mean that's sort of the closest relationship you can find. Right now it's obviously different. Right now we have both this massive deployment of 5G, which is good for us since we did the acquisition of Xcerra because of the semi test position and strength in RF. And also there's 2 massive factors in automotive with EV, right, and sort of this [ immense ] drive towards being carbon neutral. At the same time that there is automated assist, they can hop in automotive. So obviously, this wasn't happening in 2010. But if we look at just purely on numbers, that was the closest correlation that I could find.

Okay. I'm just trying to get my arms around, which of these segments are more just kind of cyclical recoveries versus really true cycle-over-cycle growth drivers?

I think there's probably some recovery in automotive. But I think, as I mentioned, we're going to see multiyear growth from both automotive and mobility for the reasons we talked about, electrification, ADAS and enabled -- 5G-enabled smartphone. So projections are multiyear growth in all of those areas.

Great. Okay. That's helpful. And Jeff, I was wondering, too, when I look at your new model, how much variability is there on an EPS basis at any revenue level depending on either mix from products or customers?

I mean mix plays a -- definitely plays a role, right? Margins will vary anywhere from mid-40s to mid-50s, depending on the product. But we've obviously taken that into account as we put together the model. So I would say, yes, there's one spike in a ramp in a particular business unit that wasn't anticipated or foreseen that could influence gross margin. I would say the variability on the EPS would be relatively low at that point, but we could see some.

Okay. And then maybe a product question here. I think in the presentation, you talked about a $200 million incremental market for the inspection business driven by AI. Maybe just a little more color on how that transpires over time?

Sure, I'll let Chris answer that.

Sure. Yes. So for the inspection business, the fundamental drivers are really -- one of the main ones is miniaturization, which drives growth in some of these advanced packages like wafer scale packages that's also occurring in the automotive space. And those devices, because of miniaturization and because of higher quality standards require more inspection. And then -- and in order to get higher-quality inspection, we've extended the inspection from what you think about as visual spectrum to infrared. Infrared allows us to see cracks and micro cracks on the surface and under the surface, again, for higher-quality standards. But -- and when we implement AI algorithms, it also allows us to -- one of the big challenges we're looking at some of these defects, whether they're scratches or cracks, some of these defects don't affect performance, they're more cosmetic. And therefore, you don't want to over reject the devices and impact the yield from the customer. So the big challenge is figuring out, okay, how do you find the defects, but how do you make sure you don't over reject? And through implementation of the algorithms we use for infrared as well as these advanced algorithms with machine learning and AI, that allows us to maintain these high-quality standards while minimizing over rejects and maximizing yield. So that's kind of the color on the inspection part of this.

Okay. And is most of that market going to be driven by kind of replacement with these new technologies of existing capacity? Or is it going to be driven by new units in the marketplace that require new inspections?

It's really driven by growth, again, of these Wafer Level Chip Scale packages. So it would be adding capacity to what they already have to handle these new types of devices.

Our next question comes from Sidney Ho with Deutsche Bank.

And congrats on the strong near-term results. Maybe I'll start off with the near-term question. I promise there will be longer-term angle here. But the last update, you did not increase your financial guidance because my understanding is that you have longer lead times, you were not able to service some of these upsides. I'm curious [indiscernible] and did orders continue to increase since a month ago? Or are you able to service some of the orders upside sooner? And the longer-term angle for that question is, do you have enough capacity to meet any kind of rush orders? And maybe what kind of revenue level can you support without major capacity addition?

Maybe I'll start with the last one, and then we'll go backwards from there. So Sidney, we have capacity now to achieve the target model even beyond it. Now recall that we do have a manufacturing model that is a bit of a hybrid. We've got in-house manufacturing for contactors. There's some outsourcing in that, but mainly in-house. We've got a final assembly and final test for handlers in-house. And then we outsource 100%, our testers. So in each of those areas, we have additional capacity to take us to that target model of revenue level. So we're good there. With respect -- Luis, did you want to?

Yes. No, I'm just going to add, obviously, there is a CapEx component to this in the contactor manufacturing side. But you're right, for the bulk of it, it's [ there ]. I'm just going to say with respect to the quarter, I mean there are 2 things. So the recurring business has a shorter lead time, right? And the second one is, in our forecast, we always have a component of new equipment qualification or qualification on new customer and when does that really going to get the qualification signature and approval and revenue recognition. So a few of those actually have come in faster than we had anticipated for fourth quarter. So increase our confidence in the fourth quarter. I think it's sort of positive. I mean the customers have the demand, and they want to get this through and continue with volume introduction of these new products.

Sidney, I think you had one more question. Can you remind us?

No, I think you actually address those two questions already. So maybe moving on to a different set of questions. If I look at the mobility side, obviously, that's the largest end market segment that you have. And you have product that cut across semi test, the handlers and contactors. And if I eyeball and mix adjusted that, it looks like you guys are looking at somewhere around 10% to 15% target revenue CAGR for that segment. First, hopefully, the math is right. But the question is, how much of that growth do you think is coming from market expansion itself? And how much of that is because of share gains and SAM expansion? And maybe one more question related to this, how much of that revenue growth comes from products that you already -- you have already introduced versus there some future products that you will still need to -- will come up with?

Okay. Yes. You're very close to the growth rate that we talked about for our tester business, anyway, of 17% annual growth. So I'm going to turn it over, though, to Ian and let him dig into some of those answers for you.

Sidney, this is Ian. So let me touch on the last question first. In the next 3 years, a fair amount of that growth is centered around products that we already have or, as I described, we just introduced over the course of this past year. We are certainly continuing to invest heavily to augment those products, but -- and to also target new types of devices out in the next 2, 3 years and beyond. As I look at the second question, which is how much is market expansion and how much is share gain. I would say it's roughly half and half, maybe a little bit more tilted to market expansion. And there's 2 elements of market expansion. One is, obviously, there's an increased number of 5G phones, Wi-Fi 6, other things that were especially strong in going into the marketplace. But also, there's an increased amount of test intensity. So each one of those devices, there's more frequencies, more test time and more expensive equipment that has to go in. The second piece is share gain. So as I look at our growth this year, a lot of the share gain has come from customers who were using in-house RF rack and stack systems. And as their needs grew and they had to move to next-generation systems, they selected our platform because it's got a leading position, especially for RFICs.

There's also an element of growth in this in the contactor side. I mean it's really big on the tester and the contactor side. So Ian talked a lot about the tester side, but there's a big component of contactors. I don't know if Devin, you want to tee off some of that as well?

Yes. And what I would say is that we do have an existing decent business with mobility. And so a lot of our growth is going to occur due to going into the higher frequency ranges, right? So we've got products that are there. We already are there with products. We are enhancing those products. And so those will help drive that share gain for us.

Great. Maybe one last question for me. Luis, in the past, you talked about this cross-selling opportunities in 2021. And I think you talked about $35 million. I assume you already included those revenue growth opportunity in your midterm forecast. But can you talk about -- give us an update what the opportunity looks like beyond 2021? Where are you seeing the best traction? And in general, how do those engagements with customers look like?

So if we start from today, right, we have a handful of customer engagements. A couple that have already converted and selected our platform, but volumes are expected beginning of the year and for the balance of next year. And you're correct. Those are included in the midterm and particularly in 2021 here. Of the $35 million, there are really 3 pieces to this equation. There are about $30 million or a little less than that, that are the equipment side, the testers, the handlers and the balance of it is the contactor. As we model it for the following years, it's really included in the growth rate that we were talking. So it's a subset of the growth rate. So when we talk about growing the tester business at 17% or the contactor business at 20%, it's really all a subcomponent of that. And you can -- as you can imagine, on the contactor, it's a very large component because the contactor strategy is really associated with selling into our installed base, so the cross-selling opportunity is the big element for contactor growth. I'm not going to be able to break it down for you how much in 2022 or 2023 right now. I don't want to start peeling it off beyond in -- that deep. But you can imagine that like I said, majority of the contactor growth is cross-selling, but there are elements of it in the tester and the handler projections as well.

Our next question comes from Craig Ellis with B. Riley Securities.

Jeff, I just wanted to start with a clarification to make sure I understood what you're seeing in the business as you revised guidance last night into today's event. So I think what I saw in the release and what I've heard today is that significantly, it's on the mobility side with some share gain in testers, but also with automotive. Were any of the other end markets like industrial or compute contributing? And as you look to the color you provided for the calendar first quarter being up 5%, is that just a continuation of the strength we're seeing in auto mobility? Or do you expect, at that time, we'll see some different gives and takes in the other end markets, just a further detailed view of the current and first quarter environment would be helpful.

Craig, I would say, yes, you're largely correct in that it's -- the story in Q4 is largely driven by RF test and automotive. Industrial, I think we see a small increase in industrial. But by and large, what's driving the business at the moment is related to mobility, RFIC test as well as automotive. And that story continues into Q1.

I'm just going to add to it. I do see a trend that industrial will be improving in the first quarter. I think that is rising right behind automotive. I think it's coming strong behind it.

And what about compute data center then, Luis? We heard, I think it was the week of Thanksgiving from Analog Devices that they have seen industrial certainly broadening out. So if you see improvement in the first quarter, that would make sense. But Micron, when they guided or talked about cloud being much more stable and they're seeing some increased strength in their business. So is compute data center starting to come back? Or is that something that might come back later in calendar '21 for the company?

I think it'll be later in calendar '21, Craig. It hasn't necessarily gone weak. It's been a healthy market this year, perhaps stronger in the first half than in the second half. But hasn't seen any cyclical down. It's more of steady as she goes. We think there will be another upside opportunity in the second half of next year, but nothing visible at this time.

Okay. Got it. And then a follow-up to one of the issues that Sidney honed in on and it's regarding share gains. So it sounds like the transition to 5G is catalyzing customers to go from internal solutions to Cohu. The question is, are there other transitions coming across other end markets that you see that could provide similar such share gain inflections as we look out over the next 2 to 3 years?

Well, I mean, we could expand and describe more detail what's going on in the RF space because I think that is the biggest wave of transition into the Cohu platforms. There's definitely expansion in automotive, but that is an area that we already have a -- bought a 60% share in the handler automotive market. And there's a new technology coming up, which is beneficial to us. But perhaps maybe we pass it on to Ian and have Ian depict a little bit more what's happening in the RF space. We tend to talk a lot about mobility and 5G, but there are other standards and other opportunities in our app that we're broadening, which I think is beneficial to our platform, including structural tests. So maybe, Ian, you can expand that a little bit for us?

Yes. I'd love to do that. So Craig, let's just start with RF and think about RF outside of mobility. So some of the new technologies coming along that really lend itself to our leading position in testing RF front-end ICs are also moving into other markets. So as an example, satellite Internet communications vehicle-to-vehicle are all seeing a pick up of technologies that are looking for more bandwidth, lower latency, all the same things that are driving the increased test intensity and the increase in these ROCs in 5G. So what that means for us is we're seeing more adoption, more design wins in market with customers outside of mobility for RF, and our existing customers who are obviously very strong in mobility are also increasing their attach rates into automotive and on the other places. And we've seen some of the share gains this year, but we think this is going to be sort of waves of activity that are going on over the next number of years, sort of alongside 5G and continuing beyond it. Now I did want to touch on when you talk about share gains in other markets. So this gets back to our core strategy in test. So we've picked, focused end applications and markets that really align to our air cool platform that we think give us and our customers advantage versus other platforms and other competitors in the marketplace. And some of these, we have relatively low share today. Flat panel display drivers, as an example, which is in mobility, but also in computing; power management, which extends across, especially into automotive and industrial. So all of these are areas that, as part of our 17% growth plan, we're expecting to drive primarily share gain in order to increase our position outside of RF and sort of leveraging our universal platform that can test RF power management, structural tests and so on. So I hope that helps a lot.

That helps a lot. Jeff, if I could just jump into some target model question. So I just want to understand the contrast between the intermediate term models, $940 million and $3.60 in earnings. And by the way, congrats on all the incremental leverage with the model of 20% is pretty remarkable. And that, with the high end of the quarterly model which annualizes at [ $1.4 billion ], what's the difference between the high end of the quarterly model and intermediate-term models, $940 million, $3.60?

Well, we thought it would be appropriate since there is some seasonality to the business that we not only had revenue targets, quarterly revenue targets, that led up to the 2 35, but also went beyond. So it's -- as we're all aware, seasonality provides different revenue rates per quarter. So we thought it was appropriate that not only did we show a lead up to that target, but also what the profitability is beyond it.

Okay. And then switching to the free cash flow part of the new target model, so $160 million, quite impressive. The model also expresses potential debt paydown of about $6 million annually. Clearly, you're on a pace that's far greater than that in the current quarter. So is the model intended to give you the flexibility to be either consistently active with small tuck-ins with M&A? Or if you couldn't execute either tuck-ins or more transformative larger deals as we saw with Xcerra, would you just take the debt paydown portion of the model up above $60 million annually?

Yes, that's right, Craig. It does give us that flexibility. We'll continue to monitor it. And with the priority for cash being to grow the business, support the business, but also to deleverage. So yes, yes, you're correct. If there were no other opportunities for SAM expansion or technology tuck-in, yes, we would have more meaningful debt repayment.

Got it. And then the follow-up is for you, Luis, and I'll use this as my wrap up and get back in the queue. Can you just talk about where you might have interest in M&A? I think you talked about potentially doing things that either expanded the SAM or tech tuck-ins. Where do you feel like you've got an opportunity to bolster your current technology base? And where might you be seeing a new SAM opening up where you'd want to have first-mover advantage or an opportunity to get in on something as it starts to inflect?

Sure, Craig. Well, we are a test and inspection company. So I think here, if you just hold that for a second and for definition, means opportunities that will help expand our portfolio, accelerate our growth in these high-growth end market segments we described today, whether it's semiconductor test or an inspection of the semiconductors or things that help optimize the overall efficiency of the test cell, help us deliver greater value through the test cell or also help us accelerate time to yield or improve yields at our customers. It could be like Chris described, the infrared technologies is one piece, the active thermal control is another one, the precision measurement instrumentation. So again, areas that add up to our strategy here to get to $940 million in higher profitability faster is what we're interested in.

And given the crazy year that we've had in calendar '20, where does the deal funnel stand right now? Do you have an active M&A funnel? Or is that something that you plan to work on as we head into calendar '21?

We always have an activity around M&A, Craig. Even after we did the Xcerra acquisition, obviously, we were in no position to go out and do another one. But we always have an active engagement and portfolio of opportunities that we monitor. It's part of our process.

Our next question comes from Christian Schwab with Craig-Hallum Capital.

Congratulations on the strong results, guys. Ian, can -- in the semitest share gains that you talked about and the shift from in-house rack and stack to Cohu, can you give us an idea if that's a couple of substantial customers who made that decision? Or is there many customers in their share gain thoughts?

Yes. I mean, we don't share a lot of information, Christian -- this is Jeff -- on customer specifics. I think we have talked about a couple of wins, a couple of major wins, as a matter of fact, in past quarters. Ian could probably add some color to this, but it is -- it's more or less a broad-based customer base.

Yes. So as Jeff said, I don't want to point out specific customers, but it is an interesting trend that we've seen, not only in customers that are new to Cohu, meaning they hadn't previously used Cohu's test equipment, and we certainly have some significant ones there. But we've also seen this in well-established customers, customers who've already adopted our platform a number of years ago, where they had a mixed strategy of in-house testers and our testers. So we've really seen this almost across the board in new customers and existing customers, which I think is a strong testament to sort of this transition, the industry going on for a need for moving to higher production-worthy systems, but also sort of the investment we've put in place to really extend our capability into 5G and Wi-Fi 6, a lot of these new standards that customer need to move to a high volume.

Okay. And then, not that 18% revenue growth isn't quite strong in the mobility side, Ian, on the test side again, but we do have 5G phones that are going to begin to ramp materially over the next 2 to 4 years, which have 30%-plus more RF content in each phone chip versus a 4G phone. And they also have 30% to 40% increased test intensity time to test those facility chips for 5G. So if we double smartphones this year, and maybe we'll double them again afterwards or maybe it slows down depending upon whose number you want to believe or how many 5G phones are actually going to be shipped this year, but let's go with Qualcomm's number of 750 million by 2022. That looks [indiscernible] with the market share gains that you've talked about or not?

Sorry, Christian, you cut out just for a moment there. Can you repeat that question? Just the very end.

I was just wondering if that 18% growth rate look conservative given all the data that I just talked about. The doubling of phone growth, we're probably going to have smartphone unit growth in aggregate again next year on top of that versus this year, more RF content, more test intensity. That seems like that could be a conservative number over the next few years.

Well, I think we're certainly hoping that that's the case. But maybe, Ian, you want to give a little more color on that growth rate?

Yes. Sure, Christian. So certainly, we're looking at this increase in number of cellphones as one of the key drivers. And also, as millimeter wave comes on board, it also -- there's different models of how much adoption rate we'll see over the next 3 years on millimeter wave, but that again increases sort of the test intensity. Now one of the things we focus very hard on, and it's a key value we're delivering to our customers, is we are constantly improving our test instrumentation and our software for existing instrumentation to reduce the overall test time as these new standards come on board. So to some degree, to be a leader in this market, we have to be sort of improving on the number of devices that can be tested in a particular test cell. So that is why it's not just a sort of a mathematical number of taking the number of new phones coming out every year and saying, what is the end market growing? So that's one of the dampening factors. But I will say, we try to have not a conservative number or not an overly optimistic number, but what we think is a realistic number for the sort of rapid growth in the market and how much share we're going to continue to grow within it.

Great. And then I guess my last question, a follow-up. I don't remember who asked it. But talking about not just seasonality quarter-to-quarter, but cycle seasonality, peaks and troughs. I don't ever remember there being a time frame. I know we mentioned 2010 somehow or another. But I don't remember ever seen a cycle where you have multi-year tailwinds that you have dominant market share. We haven't had a refreshed phone cycle with more RF content where you are well positioned across the test cell. And we certainly haven't seen the likes of which could be extremely strong automotive growth for the next decade. Am I thinking about that right?

Christian, it's Luis. Yes, you're thinking about it right. That's why we referenced it as sort of a multiyear megatrend here or a series of megatrends because you're right, we have a perfect storm in automotive and mobility with big technology transitions happening in both segments.

Our next question comes from Quinn Bolton with Needham & Company.

I wanted to follow-up on Brian and Christian's questions here on sort of the sustainability. I guess when I look at your calendar '21 -- sorry, calendar '20 revenue of $631 million, you guys, in the first quarter, per your guidance, are probably going to be run rating at close to $820 million to $840 million. And so you've clearly seen a large step function up towards your new $940 million target. And I guess my question is, do you think things in the fourth quarter and in the first quarter are a little overheated now? Or do you think we're just really catching up for a 2-year -- from a 2-year downturn in at least the automotive market and from the level of kind of Q4, Q1, the trends in automotive towards ADAS and XEVs and the trends toward 5G and RF, kind of now provide a new baseline of growth with perhaps some seasonality along that new baseline?

Quinn, it's Jeff. We look at it more towards your latter explanation. Yes, there's a bit of catch up here at -- near the end of the year for automotive, but we had a second half pick up in mobility. So what is driving us into 2021 are really these tailwinds that we've been talking about. In terms of sustainability, again, we would look to these as multiyear megatrend. And so we think that this is more than just a -- well, it's hard to predict, right, out beyond a quarter or 2. But we do know that the -- what predictions are in terms of growth and it is multiyear.

And I guess, you guys historically haven't forecasted test cell utilization, you often report it historically. But I think you've said is test cell utilization of your customer base gets into the high 70s to low 80s. That's where they really start to order. And I think you said in October, it was up at 84%. Do you think the capacity that's being added in Q4, Q1 will significantly reduce that utilization rate into kind of back to the mid-70s? Or do you think it's almost catch up demand and test cell utilization is going to remain at a fairly healthy level into '21?

Again, a bit of a prediction here, but I would, again, go with your latter comment in that we would anticipate that the utilization rate remains healthy for the foreseeable future.

Yes. I think if we are running at 90% utilization, I would be concerned that there is a trend down eventually, but that's not where we're at. We're at about 84% now. It's still sort of a healthy point, and the business can -- or the industry, I should say, can operate at the low 80s consistently. So I'm not too worried. We don't forecast test cell utilization, like you said, but not too worried about it from the numbers that we have today.

Got it. And then for Devin, on the contactor side, I guess I just wanted to come back and as I think about contactors, I guess we'd sort of thought contactor growth, in aggregate, would tend to trend with IC unit growth. And one, just wanted to see, is that a good base assumption? I know you mentioned you thought the contactor market would grow about 10% for the next 3 years, which I would likely put above IC unit growth in that time frame. So just wondering, does the contactor market typically grow with units? Or are there other factors that allow it to grow faster?

Yes. I think there -- this is Jeff, Quinn. Other factors play into it. Maybe, Devin, you can give us some color on that.

Yes. I think similar to what Ian stated in like the RF market, the increase in complexity there in some of these markets is going to increase the demands and the opportunities for socket. So where that's certainly in the RF, as we've discussed. But it also is in -- and even into the computer networking with some of the key inflections.

Okay. And so that increased complexity, is that just sort of kind of longer test times may need more contactors to test a given number of units? Or does that increase complexity kind of result in higher ASPs for the contactors? Or maybe both?

Devin, what do you think about that?

Yes. No. I would say it's -- a lot of it is going to be in units for the contactors themselves, and some of that is -- a smaller portion of that is in ASP. It's getting more complex. If I go to the RF market, right, and there really are very high level of challenges when you get into those higher frequencies. And so that obviously is creating a challenge. You have to innovate around that. There's some differentiation that goes in there, and that drives a price differential associated with that. So there certainly -- the test complexity that comes up in some of these markets, there's very high complexity coming, and those -- that would certainly be a portion of it is to increase the ASP price.

Great. And then my last question for Ian, and I apologize if I missed this, but you talked about a $400 million RF test TAM. I think most of that is for mobility or 5G. I guess I just -- as we think about IoT use cases, whether it's tracking devices, whether it's factory automation, it would seem to me that there's a lot of wireless connectivity, Bluetooth, low energy, Wi-Fi going into those devices. Does that become a growing part of the RF TAM? Or do you think 5G handsets is really going to be the dominant part of your RF test opportunity for the next few years?

Ian, can you clarify that for us?

Sure. No problem. So I think coming back to the term I tried to use before, Quinn, which is sort of waves of growth. So certainly, 5G is one of the predominant waves of growth that's taking us -- our prediction for the served market that we're really focused on, from $200 million in 2020 to our estimate of $400 million in 2023. So 5G is one of those. I will say that Wi-Fi 6 is a very clear driver this year and into the coming years as that moves into not only handsets but consumer devices, as we think about them as what goes on consumer premises for -- and also into computing. But then as we look out even just a little beyond that into '21, '22, '23, we do think that satellite Internet communications, where we have a very, very strong position and ultrawide band, which is just at its initial adoption stage, which has maybe more of a place in automotive and consumer and devices, we think all of those are going to add incremental growth over that course of the time. I don't think I have it in front of me broken out exactly how much of that $200 million is just mobility and 5G, but I think it is -- I think all of them contribute to that number.

Our next question comes from David Duley with Steelhead.

I echo the comments of everyone preceding me, it's a great presentation, it's great to have all the incremental information. Along those lines, just a clarification. If you hit your numbers in the fourth quarter that you just guided to, how big will your test business be and your handler business be in 2020 so that we can gauge these growth rates that you've given us?

So the handler business -- and Dave, we do give a pie chart in the slide presentation. The handler business is roughly about 50% of 2020. The tester business is about 20 -- I believe, it's 27% or 28% of the business in 2020.

Okay. And if you -- again, if you hit your numbers in the fourth quarter, will the automotive business have filled in the, I think, it was $30 million or $35 million that was kind of dragging on results on a quarterly basis. If you hit the guidance in Q4, will you be -- will you have caught up on all of that automotive business? Or is there still more catch up to come?

From an order perspective, yes, we've caught up. But hitting those revenue numbers, we have begun to close the gap, but haven't closed it entirely.

And just for reference, it was $25 million to $30 million.

Okay. So on an order -- just to clarify, on an order basis, yes, you've seen the catch-up, but the revenue hasn't yet caught up. So that's more like Q1 and Q2 is when you should see the fill in on the revenue side from the orders you're getting in automotive now?

Yes, Dave, I think that's reasonable.

Okay. Then just the final thing for me is, you talked about this $400 million of incremental TAM in the mobility space from 5G and an assortment of other things. How far along are we at realizing that? Is there still -- how much more of that is to come? Or how much have you already achieved?

Yes. Maybe Ian can help us with that.

Yes, no problem. So let me just try to clarify sort of the numbers we provided, which is as we look at the RF served market within just the tester side, we think that is growing from $200 million in 2020 to $400 million in '23. And again, back to -- we think a very significant portion of that is -- growth is 5G. But that market, as we think about RF, which covers all the markets we serve, certainly, RF IoT, which extends into computing and industrial is a significant portion of that growth as well.

Okay. And then I do have one more question. I'm fascinated about this market share gains that you're referring to in the RF test space. I guess I'm kind of curious as you talked about, it's a broad set of customers, and we can kind of all guess who those guys are. But historically, you've done mostly power amplifiers. But I guess the share gain opportunities is you moving into other parts like switches and LNAs. Could you just elaborate a little bit more on that? And -- because it would seem like if you've moved from power amplifiers to other components from these large customers that you already serve, but the mobility growth rate could be a little bit conservative.

Ian, would you like to add some color to that?

Sure, absolutely. So I think Xcerra, historically, focused on what we call the PA market. And the PA market is a portion of the RF front-end IC market. That is the core part of our served market on the RF side. As we -- as in the last couple of years, we focused our sights on a larger market. We really said, we saw the opportunity of looking at the entire RFIC market as our primary market that we're going after on the RF side. Now what does that include? That includes PAs, which kind of roughly makes up 50% of that market in 2020. It includes low LNAs, filters, antenna modules, antenna tuners and front-end modules, which is the -- a device that has a number of these elements all put together within it. As we look at that market and also included how much of that market is served by a rack-and-stack equipment, we sort of said, okay, this is a larger market, of which we have a real opportunity of growing share in. And that's the $200 million market I described. Again, mostly in mobility, but extending outside of mobility. So that's one piece of it. You asked again about are we being overly conservative. Our estimate is we're in the market, we're roughly 40% share of that market now. And our goal is to continue to grow share within it. We also see it as a market that is attractive to a number of our competitors. And also expect to see competition from other competitors in the marketplace. So that's why we've sort of said, I think the numbers we provided are realistic, attainable, achievable numbers given the strength we have, but also includes a fair amount of share gain going forward with it.

Just to clarify that 40% market share number you just talked about, what was that of again?

That's that 40% in the RF front-end IC market.

Okay. And so your share in power amplifiers would obviously be much higher, right? Is it more like a [ 65% ] of the...

Our share of RF is higher, right. And we have -- we believe we have both more opportunity to gain share outside of that. And we've seen some of that this year. But also, we see that, with these more complex RF technologies like 5G, like satellite Internet, there is actually more growth outside of PA into some of these other elements in the RF front-end IC chain.

All right. And just to clarify these market share numbers. So 40% of the total RF front-end business, but 65% in power amplifiers, that would imply that the LNA antenna tuners and auto fronting modules are at a very low market share, perhaps in the 20s or something like that. So there's a lot of room to grow market share in these new parts that you're starting to pick up test slots for?

Yes. Yes. Again, to emphasize, I think there's opportunity for us to grow share because we're at a smaller position, but also there's also more activity and more growth there. And that's also fueling some of the folks who are deciding to switch from in-house rack-and-stack to a more high-volume production tester. And again, I'll emphasize the advantage we're providing is that for these customers who often have all of these devices in their portfolio, they can have one test platform that can test them all. So they can manage sort of the ups and downs in their quarterly business more effectively.

Our next question is a follow-up from Brian Chin with Stifel.

I know we've moved into over time, so thanks for the follow-ups here. Maybe just -- I found your slide, I think it's Slide 61, pretty instructive. Kind of share how certain key applications align to Cohu's opportunity and product offering. And I do want to hone in on sort of the middle 2 columns tied to EV and ADAS. I'm just curious, those 2 applications really seem to be driving some of the initial growth here in auto that you've seen the recent quarter moving forward. I'm kind of curious, obviously, vehicle sales in aggregate haven't bounced back yet. But these 2 applications themes are certainly incremental. So I'm just wondering, from a -- to peer into sort of your customers' test floors, is the dynamic here where their IDMs and unlike sometimes, you maybe kind of have more fungibility or reconstitution of test cells around different product sets? Are these like ED, electrification, high-voltage and ADAS, are they more greenfield test cells that you're seeing? And is that sort of the reason why the aggregate market, combustion cars obviously hasn't fully bounced back at this point to prior kind of production levels, but you are seeing the incremental out of those businesses and could continue moving forward. I'm just kind of curious if that's the dynamic that makes sense.

Maybe we should answer this in pieces here. First of all, when we talk about a test cell, we're really talking about entailing 2 -- at least 2 out of the 3 components that we provide, either a handler of a tester or with the contactors. And for these 2 middle blocks here, I think the primary discussion for us is around contactor attachment rate to the handlers. But maybe answering to your question, is it greenfield? Is it sort of a current application? Maybe I'm going to pass it on to Chris to expand a little bit more on the handler side, and we can top off with Devin, giving you some more color on the contactor side.

Yes. So Brian, this is Chris. Let's start with ADAS. So for example, the real application here, the growth of these processors, which require -- there's a lot going on in the processor where they're processing all these different information from the vehicle and they're -- the automotive guys are very cost driven. So they're looking at how to reduce package cost. And a lot of these packages now are generating a lot of heat. So the heat has to be removed. So what we're seeing is some upgrades because we can upgrade existing systems on the test floor, but the majority of the business we've seen so far is new systems, which would, I think, would imply more of a, what you would call the greenfield opportunity. And similar on electrification, it's an upgrade to a current system, but most of the activity we are seeing right now is people wanting additional system capacity.

Got it. That makes sense. And maybe that plays in the fact that these are expected to be durable, increasing trends and penetration rates, right, on vehicle sales moving forward, right? That's really helpful. And just kind of last thing quickly here. I think there was a reference made to that you're seeing a lot of -- on the test side, a lot of upgrades maybe tied to sort of the recent introduction of RedDragon instruments. I'm just wondering, are you seeing a pretty high skew there towards sort of upgrade -- instrument upgrades like with the RedDragon or kind of full system sales? I sort of imagine the margin is pretty robust in either direction.

Ian, you want to address that?

Yes, sure. So I think we're seeing a combination in this past 6 months as we're going through this ramp of both new test systems as well as upgrades. And our upgrade business is still pretty early on. We introduced the RedDragon fully into the marketplace in calendar, our Q3, of this year. And we're still at the end in Q3. We're still at only 10% of the installed base that we were able -- that have demanded upgrades from us. We do see that continuing through the course of 2021. Our target right now for end of 2021, which we think is quite achievable, is 30% of the installed base. We still see that our revenue overall this quarter and into the next quarters are still going to be higher for new systems. But any -- but quarter-to-quarter, it's going to be some mix of the 2.

Thank you. And I'm currently showing no further questions at this time. I'd like to turn the call back over to Jeff Jones, Chief Financial Officer, for closing remarks.

Thank you. And before we sign off, I'd like to let you know that Cohu will be participating in a number virtual investor conferences over the next 6 weeks, and we'd welcome meeting with you. The conferences are the D.A. Davidson Investor Conference on December 15, the 12th Annual CEO Summit on December 16, and the Needham Conference on January 12 and 13. So on behalf of the entire Cohu team, I'd like to thank you for participating in today's event. And we look forward to meeting with you again soon. Thank you, and goodbye.

Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.