This conference call. At this time, all participants are in a listen-only mode. After the speaker presentation, there will be a question-and-answer session. To ask a question during the session, you will need to press star one one on your telephone. I would now like to hand the call over to Andy Blanchard. Please go ahead.
Thank you, Latif. Good morning, everyone, and welcome to our discussion of Teradyne's most recent financial results. I'm joined this morning by our CEO, Mark Jagiela, President Greg Smith, and CFO Sanjay Mehta. Following our opening remarks, we'll provide details of our performance for 2022's second quarter, along with our outlook for the third quarter of 2022. Press release containing our second quarter results was issued last evening. We're providing slides on the investor page of the website that may be helpful to you in following the discussion. Replays of this call will be available via the same page after the call ends. The matters that we discuss today will include forward-looking statements that involve risk factors that could cause Teradyne's results to differ materially from management's current expectations.
We encourage you to review the Safe Harbor statement contained in the earnings release, as well as our most recent SEC filings. Additionally, those forward-looking statements are made as of today, and we take no obligation to update them as a result of developments occurring after this call. During today's call, we'll make reference to non-GAAP financial measures. We've posted additional information concerning these non-GAAP financial measures, including reconciliation to the most directly comparable GAAP financial measure where available on the investor page of the website. Looking ahead between now and our next earnings call, Teradyne expects to participate in technology or industrial-focused investor conferences hosted by KeyBanc, D.A. Davidson, Jefferies, Deutsche Bank, Citi, Evercore, Piper Sandler, and Goldman Sachs. Now let's get on with the rest of the agenda. First, Mark and Greg will comment on our recent results and the market conditions as we enter the new quarter.
Sanjay will then offer more details on our quarterly results along with our guidance for the third quarter. We'll then answer your questions, and this call is scheduled for one hour. Mark?
Hello, everyone, and thanks for joining us. Greg will summarize our Q2 results, and I'll comment on the full year outlook and technology drivers we expect in 2023. Sanjay will then take you through the financial details, including our outlook for the third quarter. Greg?
Thanks, Mark, and good morning, everyone. Teradyne's second quarter sales and profits were above the midpoint of our guidance in revenue and profit, with sales of $841 million and $1.21 in non-GAAP earnings per share. From a market perspective, we're seeing mixed signals. While the pressure to improve lead times in some test markets is as strong as ever and we continue to work through supply issues, we have seen some softening in mobility demand, our largest test market. Also, it now looks like IA growth in the second half will be similar to first-half growth. In today's call, we'll try to provide some context on how we're navigating this complex environment. Diving into the details of the second quarter, semiconductor test revenues were in line with our plan, as were profits.
The automotive and industrial end markets served by our Eagle platform were notably strong in the quarter. In memory, flash final test and DRAM wafer sort were the strongest markets. From a strategic perspective, Semi Test had a strong quarter of competitive design wins for both R&D applications and for devices that we expect to ramp in volume production over the next year or so. The design wins were in a wide range of end markets. There were UltraFLEX family design wins in mobility, where the system's architecture supports higher throughput for large, complex devices, especially when time to market is an important factor. Eagle test had a significant design win in a battery management application, where the system's highly accurate voltage measurement capability enables customers to get to market faster with better device specifications, delivering higher value.
In systems test, sales were on plan in the quarter, with solid year-on-year growth for storage test and production board test. At LitePoint, strong wireless test demand for Wi-Fi 6E, Wi-Fi 7 R&D, and Ultra-Wideband test products drove double-digit year-on-year growth. Notably, we're seeing strong interest in Wi-Fi 7 development test, where we are qualified at all major chipset vendors, and to date, we have won 20 of 21 design-in opportunities with end product design teams. This sets us up to expand our already high share of Wi-Fi production test as these products ramp. In industrial automation, revenues grew 10% from last year's Q2 and 19% for the first half. This lower than planned growth in the first half is attributed to two primary factors and warrants a few comments. The first factor was foreign exchange. The majority of IA revenue is linked to the euro.
For example, within Europe, at UR, we saw a robust 34% unit growth in the first half of 2022, but the decline of the euro with respect to the dollar has muted UR revenue growth in that region to 20%. The dollar's appreciation was a revenue headwind in the first half, and Sanjay will comment further on the full year foreign exchange outlook. The second factor is lower regional demand that can be traced to a variety of region-specific causes, including COVID lockdowns in China and distribution partner staff shortages in North America, where end demand remains quite strong. Staff shortages for integration partners is an ongoing challenge.
In the past, we've noted that our distribution partners struggle to grow staff at the same rate as IA demand growth. Our strategy to overcome this is to use software technology and plug-and-play applications from UR+ and MiRGo partners to shorten the deployment time to make these critical staff more productive. An equally important part of the strategy is to build a robust OEM channel. For example, at Universal Robots, our OEM channel, which includes verticals like welding, power grid maintenance, and order fulfillment, is doing very well, growing 39% in the first half of 2022. The OEM channel is an important complement to our distributor-based channel because these OEM partners deliver a mostly complete solution to their customers that can be put in operation almost immediately. In many cases, these OEMs have an existing distribution network that can reach customers beyond our traditional channel.
These partners can scale much more efficiently than distributors that add staff in proportion to revenue growth for customer support. We are actively adding new OEM partners to serve these and new market verticals. Now, I'll hand it back to Mark for more on the market, the second half outlook, and the demand drivers for next year.
Thanks, Greg. While the long-term drivers in our test and automation markets remain firmly in place, the recent deceleration we're experiencing in test will impact our full year results. Over the past few weeks, we've seen a market slowdown in SOC and wireless tester demand in the second half related to declines in end market shipments of smartphones, compute products, and associated infrastructure. These demand adjustments are happening in real time, and we are projecting this will continue over the next several weeks as the market continues to align their production plans and inventory levels to this new reality. This extends across multiple customers and device types, including apps, processors, power management, and RF. I will note that auto, MCU, industrial, and memory test demand remains strong entering Q3, and we've seen no deceleration of demand in these markets.
When forecasting the second half of 2022, we are expecting the test realignments we've seen in the past few weeks are not a blip and that further adjustments are coming. Beyond the reductions we've already seen, several customers have indicated they are in a capital review phase, which we're assuming will lead to some additional impact to Q4 demand. At a macro level, we also assume that China smartphone volumes will not accelerate through the remainder of the year. We also believe that short-term global economic conditions will remain weak, resulting in softening consumer demand and associated electronics inventory digestion. In IA, we forecast that our second half business will continue to grow at about a 20% year-over-year level, consistent with our achievements in the first half, but below our 35% goal.
As Greg highlighted, we believe that China demand will remain muted, distributor labor shortages in North America will limit the rate of installation expansion in the short term, and FX headwinds will not abate. Summing it all up, we now expect our second half revenue to be slightly lower than the first half, resulting in a roughly 52%, 48% first half, second half split. Our latest forecast projects the 2022 SOC test market will be in the range of $4.2 billion-$4.6 billion, down from our April estimate of about $5 billion. While in memory, we expect the market will remain at about $1 billion. To reiterate, much of our forecasted slowdown assumes continued, yet to be identified, demand decline in the fourth quarter.
Looking further ahead, after six consecutive years of robust SOC test market growth, we have entered a period of demand slowdown and digestion. This is not uncommon as we've seen this most recently in 2013 and 2015, and over the years, we've built our operating model to anticipate these kind of demand swings. Predicting the depth and duration of these corrections is challenging, but in each of the last three corrections, growth returned after six to 12 months, with the subsequent years showing strong growth. The ramp of 3 nm starting in 2023, followed by Gate-All-Around and increasing multi-chip packaging remains unaltered drivers of growth ahead. Interface transitions in both flash and DRAM are also imminent. Hyperscalers continue to expand their chip design starts, and Edge AI is a new class of silicon application gaining momentum.
In industrial automation, the value proposition has only strengthened as the short ROI remains compelling, while increased resiliency in a world of questionable labor supply adds another motivation to automate. While the first half of 2022 unfolded roughly as outlined in January, the second half is looking considerably softer than we expected just three months ago. However, we are confident of our long-term growth strategy and are focused on execution of our design-ins and R&D programs so that we are positioned to capitalize on growth of these markets that we serve. Sanjay will now take you through the financial details. Sanjay.
Thank you, Mark. Good morning, everyone. Today, I'll provide details on our Q2 results, offer additional color on how we're addressing both the slowdown in some markets and supply shortages in others, and describe our Q3 outlook. Now to Q2. Second quarter sales were $841 million with non-GAAP EPS of $1.21. Non-GAAP gross margins were 60.2%, and our non-GAAP operating expenses were $251 million, $14 million below mid-guidance. The main driver of the lower than planned OpEx was an industrial automation group or IAG, due to lower variable compensation tied to lower revenue projections and slower than planned hiring. Non-GAAP operating profit rate was 30.3%. We had two 10% customers in the quarter.
The tax rate, excluding discrete items for the quarter, was 16.9% on a non-GAAP basis. Please note, you should now use 16.5% for the full year non-GAAP tax rate. Looking at the results from a business unit perspective. Semi Test revenue was $541 million. SOC revenue was $461 million, driven by strength in automotive and industrial markets. Memory revenue was $81 million, led by flash final test and DRAM wafer sort. System Test Group had revenue of $135 million, which was up 29% year-over-year. Storage test sales, including both HDD and system-level test solutions, were $86 million in the quarter, up 49% from Q2 2021. Defense and aerospace and production board tests combined grew 4% year-over-year.
At LitePoint, revenue of $64 million was up 16% from prior year, due primarily to strong shipments in Wi-Fi 6E and Wi-Fi 7 and UWB test systems. Now to industrial automation. Industrial automation revenue of $101 million in Q2 was up 10% year-over-year. This was lower than expected, as Greg noted. Despite the lower growth, we still expect IA revenue to follow the historical pattern and grow as we move through the year. UR sales were $83 million in Q2, up 8% year-over-year, with the highest growth in Northern Europe. MiR sales were $17 million, up 9% from Q2 2021 in the quarter.
From a financial perspective in IA, the group was slightly under break even on a non-GAAP operating basis in the second quarter, and for the full year, we expect to be towards the low end of the 5%-15% profit range we discussed in past calls. We view the roughly 20% growth rate in IA as a short-term situation, as Greg noted. Like the company model, the IAG group operating model naturally flexes spending down based on profitability, and we're tightening discretionary spending where appropriate. Our long-term IA growth strategy and related investment plans remain unchanged. Shifting to supply. Our Q2 guidance excluded approximately $50 million of revenue tied to our inability to supply customer demand. In Q3, we're excluding a similar $50 million of revenue from our guidance range, primarily in our test businesses.
The shortage of semiconductors, ranging from FPGAs to industrial analog, continues to impact our production. As I've noted in prior calls, we're taking numerous actions to harden our supply chain, but even with these actions, we expect supply line constraints to remain challenging. I will note these actions and other factors improved our supply situation in Q2, mainly in test, which enabled higher shipments. Shifting to the balance sheet and cash flow. Our cash and marketable securities at the end of the quarter totaled approximately $900 million, down from $1.2 billion at the end of Q1. We had $70 million in free cash flow in the quarter, reflecting the timing of shipments and supplier prepayments.
Over 80% of our shipments occurred in May and June, so our DSO expanded to 74 days, which we expect will decline in the second half of the year. Other uses of cash in the quarter included share buybacks of $331 million, dividend payments of $18 million, and debt retirement of $22 million. At the end of Q2, we're more 2/3 of the way through our $750 million share repurchase plan for 2022, with $217 million remaining. In our October call, we'll note any updates to our share buyback plans. Regarding debt, to date, $386 million of convertible bonds have early converted. I'd also like to note several points regarding the strong U.S. dollar and its impact on our results.
In our test portfolio, the majority of revenue and expense are in dollars, so there's not a material foreign exchange impact. In IA businesses, on the other hand, have a large amount of euro-linked expenses, and about 50% of their revenue is tied to the euro. The result is the strong dollar reduces IA's revenue and gross margin in dollars. In the first half, the FX impact reduced our IA growth rate approximately 4 points. Assuming exchange rates remain consistent with July's exchange rate, we expect 6 points of growth headwind for the full year compared to our January projection. For IA, the strengthening dollar has a marginal benefit on the OpEx side. The IA revenue and margin degradation is offset by the OpEx gain, yielding a neutral effect on our operating profit for that segment. Now to our outlook for Q3.
A combination of slowing demand and test, extending lead times due to material shortages, and reduced automation demand in Europe and China results in a lower Q3 outlook than we expected three months ago. As noted, the guidance excludes approximately $50 million of shipments due to material shortages, primarily in test. Also, the guidance assumes we won't see any extended shutdowns of production facilities due to COVID, and we won't see any new trade restrictions. With that said, sales in Q3 are expected to be between $760 million and $840 million, with non-GAAP EPS in a range of $0.90-$1.16 on 166 million diluted shares. The third quarter guidance excludes the amortization of acquired intangibles.
Third quarter gross margins are estimated at 58%-59%, down from Q2 due to product mix, 2022 investment supply chain resiliency and wage inflation. Some of these effects are transitory, and we expect our midterm earnings model gross margin range of 59%-60% to remain intact. OpEx is expected to run at 31%-34% of third quarter sales. The non-GAAP operating profit rate at the midpoint of our third quarter guidance is 26%. As Mark noted, we expect second half revenue to be below the first half, approximately 48% of full year sales. Given the reduced revenue outlook for the second half of the year, our operating expenses are now planned to grow just 4%-6% annually versus 11%-13% planned in our prior guidance. The OpEx reduction from prior guide is driven by two factors.
First, our variable compensation model is approximately half of the decline tied to reduced revenue. Second, we have delayed some expenditures in both Test and IA, which we believe will not impact our long run competitiveness. Spending savings are roughly evenly split between engineering and go-to-market. Summing it all up, we're expecting revenue and profit in Q3 in the second half of the year to be lower than we projected three months ago, but our operating model is resilient to varying revenue levels. With the reduced revenue level, our model reduces variable compensation expense, and we're taking other steps as appropriate. A downturn is always challenging, but we're confident we have the operating model, strategy, and experience to lead us through whatever lies ahead while keeping our focus on the needs of our customers and the long-term opportunities in the test and industrial automation markets.
With that, I'll turn things over to Andy.
Thanks, Sanjay. Latif would now like to take some questions, and as a reminder, please limit yourself to one question and a follow-up.
As a reminder, to ask a question, you will need to press star one one on your telephone. Please stand by while we compile the Q&A roster. Our first question comes from the line of Vivek Arya of Bank of America. Vivek Arya, please go ahead.
Thanks for taking my question. I just wanted to clarify that you're guiding, you know, Q4 roughly around $670 million-$675 million, which seems to be kind of pre-COVID levels. I think if I take your IA guidance into account, it suggests that non-IA Q4 sales would be $430 million, which would be, I think, almost 20% below pre-COVID levels. Did I get that right? I realize that there are headwinds, but, you know, why such a sharp decline in your Q4 outlook?
Yeah. Roughly speaking, those are about right. I think as I mentioned in my prepared remarks, you know, maybe we've probably reduced our second half revenue targets somewhere between $300 million-$350 million. About a third of that is known customer weakening, where the demand has been communicated to us to have softened. 2/3 of it is just a projection of more to come, in that vein based on conversations that are just beginning with customers. All of this is mostly related to Test. There's not sort of a clear identified 2/3 of that decline portion, but it's just the experience we've seen in prior cycles in the test industry suggests that there's more to come.
If you go back to 2013, for example, when the market went through a significant 30% year-over-year correction, this was sort of the beginning of that kind of trend. Now here we're talking about a market correcting about 9% off of the peak of last year. A much more modest correction. But that's. This is kind of how the market turns, and we're using history to sort of guide us in how we're projecting these yet to be identified digestion phases.
Got it. Then, Mark, I'm curious, what is your exposure to Android smartphones, you know, this year versus last year? How much of a headwind, you know, is that, as that continues? Do you think that will alleviate next year, or can that still be an offset to any 3 nm testing benefits on the iOS side? Just the interplay between Android and iOS kind of this year and, you know, what you're thinking about from a next year perspective. Thank you.
Yeah. I think our exposure to Android smartphones is a little bit higher this year than it's been in past years. We've picked up a little bit of market share in that segment. Obviously that segment is down significantly this year in unit volume, so that's principally the reason for the declines we're forecasting here in the second half. Now when we look at 2023, it's hard to tell what's gonna happen with Android phones. Part of what's happened in 2022 in terms of the declines related to China, you know, and part of this is predicting how China will rebound in 2023 as the consumer demand hopefully recovers.
We don't have a strong forecast for the Android subsegment of smartphones yet for 2023, so I can't really give you a lot of color on that.
Thank you.
Thank you. Our next question comes on the line of Toshiya Hari of Goldman Sachs. Toshiya Hari, y our line is open.
Hi, good morning. Thanks so much for taking the question. I had two as well, one on Semi Test and the other on IA. On the Semi Test side, Mark, I was hoping you could speak to your expectations in terms of market share for the year, given the incremental weakness you're seeing at the market level, and given that you had sort of de-risked your opportunity for this year with your largest customer. I feel like your market share position as of today is perhaps a little bit better than what you were thinking three months ago. I just wanted to clarify that on the Semi Test side.
I think there are two pieces of Semi Test, memory and SoC. Memory, it's gonna be 40-ish% market share consistent with last year. On SoC, it really depends on what happens on the compute side in terms of how much decline happens in the market on compute, which is more of our competitor's strong suit. We've baked in a significant decline in mobility. We believe a decline in compute is imminent. That's something that I think over the next few weeks, as we hear from our competitor, we'll know more about. All that being said, I think our statistical share is sort of in the upper mid- to upper-30% range in SoC test this year. That's based on, again, strong compute market, weak mobility market versus customer switching suppliers.
That moves around quite a bit year- to- year.
Got it. That's helpful. As my follow-up on the IA side, I guess a two-part question. The first part is I just wanted to clarify that, you know, fundamentals in the business remain pretty strong. I ask the question because the reasons you noted for, I guess the reduced outlook for the second half seemed, you know, technically transitory. You talked about FX, you talked about COVID lockdowns in China and staff shortages in the U.S. Fundamentally, is the business still healthy and intact? I guess that's number one. Number two, on the OEM channel dynamic, just curious, how big is that channel today as a percentage of IA, and where do you see that going in a couple of years?
How does that impact profitability, as OEM grows, vis-a-vis your disti business? Thank you.
Hi, this is Greg. Let me try and take that. We believe that the fundamentals are still very strong in the IA market. The effects of China lockdowns and staff shortages should both ease to some extent through the rest of the year. We believe that the FX headwind is gonna be consistent through the rest of this year. We don't see a significant appreciation of the euro, so that translation into dollars is gonna continue to take a couple of points off the top in terms of our growth for revenue that's denominated in euros. To your second question, our OEM channel right now it represents about 16% of overall UR revenue, lower for the entirety of the Industrial Automation group.
It's growing, for the first half it grew at 39%, and we expect to be able to maintain that kind of a growth rate.
Impact on profitability?
Our pricing in the OEM channel is a little bit favorable to pricing through our normal distribution channel, so we don't expect to see any negative impact to profitability as that channel grows.
Great. Thank you for the details.
Thank you. Our next question comes from the line of Samik Chatterjee of JP Morgan. Your line is open.
Thanks for taking my questions. I have a couple as well. I think, within Semi Test, you mentioned you're continuing to see robust demand in autos and industrial within Semi Test. I was just wondering, given the trends you're seeing in industrial automation, are you baking in any incremental weakness as you go through the remainder of the year in autos and industrial as well within Semi Test? I have a follow-up. Thank you.
No, not really. Most of the weakness we're baking in, again, in the fourth quarter that we've yet to really get confirmation on is more in mobility and compute and infrastructure. We think industrial and auto will remain strong through the end of the year.
Okay. Just to follow- up on the cost structure here. I see, I mean, you've obviously pared back some of your expenses that were planned for this year. In relation to sort of SG&A, you're running pretty much at the same rate that you were last year when revenues were much higher. As you start to think about sort of the strategic CapEx review by some of your customers and maybe some delays to the secular long-term drivers that you have, what is the sort of flexibility in terms of taking the cost structure lower or sort of right sizing it for maybe some of the sort of push outs in terms of revenue that you're seeing relative to just sort of postponing spend as you're doing, sort of announcing today? Thank you.
Hi, it's Sanjay. Just a couple of comments on our business model as it relates to cost structure. As you know, the majority of our test portfolio has contract manufacturing, so it really provides great scalability up and down with our outsource partners' kind of fixed assets. That helps on the margin line. From an operating expense perspective, we have a key portion of our wages in a flexible or variable compensation plan, that is, as revenue and profit goes up, that is passed along part of it to the employees. As it goes down, which is one of. It's about half of the driver of the growth reduction from a midpoint of 12% expected this year to a midpoint of 5%, as I noted in my prepared remarks.
About half of that growth reduction is driven by the variable compensation. We believe as time moves on and, let's say, the market goes lower, we'll have the ability to reduce our operating expenses on a variable basis while keeping our workforce intact, and investing for the long term. You know, obviously, as the market rebounds, or increases, our revenue goes up, then that will increase our operating expenses on a variable basis.
Thank you. Thanks for the color.
Thank you. Our next question comes from the line of C.J. Muse of Evercore ISI. Your line is open.
Yeah, good morning. Thank you for taking the question. I guess first question on gross margins. If I look at what you've kind of discussed in terms of mix, that's actually a positive on the SoC side. Yet you're guiding gross margins, I think the worst since kind of mid-2020, at a revenue run rate I think that exceeded back then. Curious, what is driving the gross margin hit here? Is it a much worse mix in SoC than what perhaps I'm contemplating? Or is it kind of the FX other issues around IA?
Sure, C.J. I think when you think of the product mix, there is some mix degradation within Semiconductor Test, but also as there's less wireless test, which is higher than the corporate average, and less than expected kind of IA, which is higher than the corporate average. You are seeing mix. A couple of other drivers that I noted in my prepared remarks, and that is we've spent a lot of effort and monies on making our supply chain resilient through component qualification, through multi-country manufacturing, et cetera. A lot of those costs are gonna be incurred in 2022 and the second half of 2022. We have those that significantly curtail.
There will be some in 2023, but we'll curtail the majority of them in 2022, so think of those as transitory. Then we're also seeing a little bit of wage inflation for both ourselves and our partners.
Very helpful. I guess as my follow-up, in terms of the slowdown that you're seeing in test, are you seeing pushouts or actual cancellations? Then as you think about 2023, obviously a lot of changes afoot in terms of heterogeneous compute, larger die sizes, smaller die sizes with chiplets, advanced packaging. Curious, you know, how you're thinking about that net effect into 2023 and how whether you know we could see growth or further declines and what your market share might look like. Thanks so much.
Thank you. Yes. On the, you know, heterogeneous computing and the effects in 2023, I think that's all positive for us. You know, 3 nm is coming as planned. We don't see any changes or delays in anything associated with that. The trends toward chiplets, multi-chip packages, and the associated test intensity increase around Known Good Die, more pins to test, you know, in a chiplet design than there would be in a monolithic design, all are tailwinds for test. Now, those have been occurring in the market and will continue to phase in over the next several years as that method of, let's say, package assembly gains more and more prevalence. That, you know, is another one of the tailwinds that gives us confidence in the mid-term earnings model that we produced early in the year.
No change to our beliefs around 2023 at this point. In terms of pushouts versus cancellations versus anything else, we're not seeing any cancellations. What we have seen is that, there's been a few pushouts, but more significantly, we do hold some capacity in place for some of our key customers on a, you know, a basically a sort of 12-16-week delivery window basis that were unbooked, and it's the demand for those unordered testers principally that disappeared.
Very helpful. Thank you.
Thank you. Our next question comes from the line of Robert Graham of Loop Capital. Your line is open.
Good morning. Thanks for taking my question, and welcome aboard, Greg. I have two, maybe more, you know, 40,000 ft theoreticals, if you don't mind. The first is, you know, when we started the year, largest customer decides to curtail tester purchases because I think as we've determined to be correct, that their complexity in this year's launch was just not enough to warrant new testers. The idea was that next year, their 2023 launch would have much more complexity aided by 3 nm. With the smartphone market weaker today, has anything changed that framework in your mind? Or more, perhaps even more importantly, if I may, what your largest customer is saying to you, Mark?
Yeah, no changes on that front. You know, our largest customer has fared pretty well this year despite overall unit declines in smartphones. It's a bright spot, if anything else, that segment of the smartphone market is robust. There's nothing we've heard that would change our view of 2023.
Very good. Then as far as this, you know, this whole six to 12 month, you know, slowdown that you've seen in the past, which is obviously documented in the numbers, Mark, in your eyes, does that actually start in 3Q because the first half was kind of more a Teradyne specific thing with that large customer? Does that essentially six to 12 month timetable start now?
Yes. I would suggest it starts in third quarter because you're right. The first half sort of market declines were pretty isolated to one customer having that big of an impact on the overall view of the market. What we're into now is something more classical, I think. It's certainly, you know, and part of this, again, I'll caution, is a projection on our part. We've seen it now in the smartphone side, but we believe and sense that it's coming in a broader set of markets like compute and infrastructure. There's inventory overhangs as well out there that, you know, typically in phases like this go through a digestion phase, and working down inventory levels take that sort of six to 12 month period.
We think that as we head into sort of lower economic growth periods and a little bit more inventory in the channel than historically has been there, that people are gonna tighten down a bit, and it's gonna slow demand for sort of repeat orders for devices in those categories. Yes, starts in Q3. Six to 12 months I think is a safe bet, and we'll see how it plays out.
Mark, if I could just sneak this one in, by extension to that, thank you for that, by the way. By extension to that, does the 2024 framework now, is that the lower end of that maybe more likely, or is the lower end in jeopardy? I'm just wondering 'cause you didn't put that framework into the deck.
No, I think. Look, I think if you look at the history of the market, that 2024 framework, our earnings model and projection is intact and in play, and I wouldn't characterize it as low end either. When you look at how much the markets move up and down year-t o -year historically, we could easily see a snapback of, you know, 20, 30% growth in a subsequent year. I wouldn't color 2024 in any way based on what we're seeing right now.
Thank you. Very helpful.
Thank you. Our next question comes from the line of Brian Chin of Stifel. Your line is open.
Good morning, thanks for letting us ask a few questions. Maybe can I just backtrack to the $600 million reduction in the SoC TAM? Can you maybe just delineate again what is mobility versus compute within that reduction? Also just sort of zooming in on the compute TAM, are you seeing or anticipating any weakness outside of processors for client PCs, you know, meaning sort of you know hyperscale or server high-performance compute chips as well?
I'll take the TAM side of it. You know, where we're seeing, in our estimates, really tied to our $4.4 billion midpoint, we're seeing about $300 million, our estimate is a $300 million decline in compute, $400 million decline in mobility, and an increase in auto of about $100 million. That's really, Brian, where we're seeing it.
Then on the question of, you know, where in the compute market are we anticipating or getting some early signals of weakness. Certainly right now, the immediate corrections are occurring related to consumer laptops, a little bit of enterprise PC area. We're not seeing at the moment the correction in cloud infrastructure or service. However, there are, I would say, discussions going on in those areas that suggest that that's probably coming. Part of what we've assumed in our both market projections and Q4 revenue projections are some declines starting in Q4 in those segments. Modest relative to mobility, however.
Okay. No, that's interesting. Maybe just switching gears then, back to automation. We'd also heard that supply issues, not even UR or MiR, but for other complementary products like 3D cameras or conveyance, had also been delaying automation projects. I'm curious if that's something you're also seeing. The kind of second part of that is, and it's not totally fair to ask, but how are you thinking about IA growth rates for next year given the existing and anticipated headwinds, including demand, not only supply?
Hi, this is Greg. In terms of supply issues, we are definitely hearing from our partners that they are having trouble procuring some elements of the solutions that they're building. I think that's part of the headwind that we're seeing in North America, that there's staff constraints, and there's also a large number of partially complete projects that they need to get off of their books. We expect as those supply constraints start to ease, that will come off as a constraint, and we think that will probably happen towards the latter part of this year.
In terms of growth projections for 2023, it's very early, but we don't see any reason to really deviate from the 30%-45% range that we've been talking about in terms of IA growth, long term, because the end market fundamentals are very, very strong. We have very low market penetration, and labor scarcity is still a big factor in all of the markets where we play.
Okay. Thanks, Greg. Appreciate the color.
Thank you. Our next question comes from the line of Sidney Ho of Deutsche Bank.
Great, thanks. I wanna double click on the mobility market. I think most of us understand the order pattern of your largest customer, but excluding that customer, is your revenue growth more closely tied to smartphone sales, meaning they are more reflective of current demand? The reason I ask is that if we start hearing smartphone inventory getting closer to normal levels, say by the end of this year, does that mean you will start seeing more tester revenue growth in Q1, or do we have to wait a little longer because your customers may not need extra tester capacity until later?
Yeah. A good question. You, I would roughly say that in the smartphone world, tester capacity precedes unit production by about three months. What we're seeing in the decline in our third quarter for tester capacity demand reflects a unit production reduction of smartphones in the fourth quarter, primarily. If we're thinking about next year, then, you know, we should start to see tester demand pick back up three months prior to major phone launches or a return to growth in the China market, for example, that's sort of the model.
Okay, that's helpful. My follow-up question is on the memory test. You called out the strength in memory test in second quarter, and you still expect the TAM to be about $1 billion this year. Many memory suppliers are likely gonna cut CapEx quite a bit and slow down production. When do you think you will feel these adjustments in your revenue, or is increased complexity and market share gains enough to offset these headwinds? Thanks.
Yeah. I think in memory it's a little different, and we don't expect to see much of a correction because there's some tester obsolescence occurring in the new DDR5 and LPDDR5 interfaces that are growing in share in the DRAM world require brand new testers. All the fab capacity that's been put in place in 2022 to produce that class of device that should start growing in the server world and in the phone world in 2023 will need to be facilitated with testers in the second half of this year going into next year. I think there's a you know a bit of a lag between the CapEx investment on the front end and the tester investment on the back end.
We think at least through the early part of 2023, despite what we're hearing in memory from the suppliers in terms of bit growth and everything else, that these obsolescence, technological obsolescence events will have to be solved with additional tester capacity because of the old tester just can't be reused.
Okay. Thank you.
Thank you. Our next question comes from the line of Joseph Moore of Morgan Stanley.
Great. Thank you. In the past, you guys have sort of talked about the relationship between the wafer fab equipment market, particularly the logic portion, and your SoC test business. You know, obviously you're kind of diverging now and you know, I'm not asking you to make a call on WFE, which I think most of us do expect will probably contract, but can you just talk a little bit about that relationship? You know, this contraction that you're seeing seems a little bit at odds with how much logic capacity is coming online and how much stuff would presumably need to be tested. Just give us some context on that relationship.
You're right. It is at odds. It doesn't correlate well. It suggests that there would be some underutilized fab capacity for a period of time or there will be wafer die banking going on. In other words, the fabs will keep running, but the devices themselves may sit untested, undiced, unpackaged for some period of time. I think that is likely to be the case for a little bit of a period of time coming in the next six to nine months. Usually, you know, historically that washes itself out at the back end of the six to 12 month correction. That's kind of, you know, what I would expect.
Okay. Thank you. Then for my follow-up, on the last question, you kind of referenced technology transitions in memory. If you talked about DDR5. I didn't hear it, but can you talk a little bit about is DDR5 timing kind of getting pushed back? Is that a factor in DRAM being a little lower this year? What does that tell you about, you know, kind of your memory business prospects for next year?
Yeah. Yeah, it absolutely does. You know, this year we expected, if you go back to the fourth quarter of last year, the DDR5 would be a bigger part of the market and our business in 2022. With the delay of Sapphire Rapids and such into 2023, a lot of that tooling pushed out. However, you know, the market kind of held up because what filled in behind that was pretty robust investment, some for new suppliers in China. You know, all year long, they've been very aggressively ramping capacity. It turns out, despite the delay in DDR5, 2022 is a pretty good year for memory. When we look into 2023, you know, there could be a little bit more of an incremental delay.
You know, I think everybody's hearing about maybe Sapphire Rapids is gonna get slid out another few months or so. It's coming. You know, we're getting close to escape velocity, I think. We're seeing the orders for capacity for that coming in now. That's what gives us a little bit more confidence on the memory side than we're seeing on the SoC side.
Great. Thank you very much.
Thank you. Our next question comes from the line of Krish Sankar of Cowen. Your line is open.
Question. I have two of them. First one, Mark or Greg, to the extent you can answer it, a question on next year. Do you worry that if and when mobile demand starts getting better next year and the auto industrial segment could roll over, and wouldn't that be more detrimental to your margins? How to think about that, you know, transition, if you can answer it, and then I had a follow-up.
Yeah. I don't think the margin between auto and mobile is that significant for us, frankly. I wouldn't say that the mix, if there's a mix shift going on there, that it's gonna be a headwind or a tailwind. They're pretty blended close to each other. As to when we would expect mobile to return, you know, that's. We know, at least from the pattern from our largest customer, that that's a 2Q, 3Q phenomena, and we see no change to that sort of timing, just based on their product introduction cycles. For Teradyne's point of view, we would expect the big changes to occur in those two periods.
There should be a little bit of pickup in Q1, modest pickup, I would say, related to some other manufacturers' phone introductions, but 2Q, 3Q is the big quarter.
Got it. That's very helpful. Then a follow-up, you touched upon the little bit of the decoupling between front-end WFE and test investments. I just wanted to like go into one subset of that, which is China. It seems like some of the Chinese OSATs have slowed down spending, but the front-end spending is still continuing. I'm just kinda curious your view on how China, you know, evolves for test spending and for Teradyne.
Yeah. You know, China is and it's not just China, but I would say that the fab investments, generally speaking, are continuing unabated. Going back to that utilization question, you could make the argument that the 3 nm investments that have been made going back now over a year are hugely underutilized because no revenue is yet to be produced off of that. So the tooling, the advanced time to tool and tune the recipe for 3 nm has become an unprecedentedly long cycle. All of that fab capacity is, quote-unquote, "underutilized" until next year revenue starts flowing. Specifically back to China, there's a disconnect between test lead times and thinking and fab lead times. Fabs tend to get put in as a whole slug of capacity, with all of the front-end equipment installed quarters before production starts.
Test tends to be added as a spot market add. Once the fab starts running, the initial capacity gets facilitated in test in a quarter. Devices ramp up and output of the fab ramps up in subsequent quarters, and incrementally, test ramps up as well. Test is a much smoother deployment than a fab. The last thing, of course, going on in China, I think, is just the national priority around independence and access to technology has focused mainly on fab technology. It's the most, I would say, viewed as sort of threatening and enabling at the same time. There's been a more, I'd say, resilient maybe counter-economical push to invest in fab capacity there. Our business in China is very strong.
We have one major headwind, which is the largest single consumer of test equipment in China, Huawei, is somebody we cannot serve. Outside of that, we have a very strong business, and a very, you know, decent growth, that we're seeing there.
Thank you very much.
Thank you. Our final question comes from Mehdi Hosseini of SIG. Your line is open.
Yes. Thanks for taking my question. I want to go back to SoC test. Mark, when I look at the trends since 2015, the SoC test, g rew on a year-over-year basis. Perhaps the two factors there, the parallel tests went away towards the end of the last decade, and also the 5G that helped. It started to be incremental in late 2019. Now we're here, you're guiding to $4.4 billion, which is still above the 2020 level. With that as context, what gives you the confidence that we should see a quick snapback in SoC test in 2023, especially since your largest customer is expected to continue with bifurcation, not moving everything to the next leading edge node? I have a follow-up.
Okay. Good, Mehdi. You're right, you know, monotonic growth since 2015 has been true. Even with, you know, the $4.4 billion current midpoint view of the market in 2022, the trend line growth off of 2015 is still kind of a 10% CAGR. We've said all along, we run the business and expect volatility and sort of manage the business to the trend line. The 10% CAGR is kind of in line with our mid-term earnings model too. It's hard after so many years of monotonic sequential growth maybe to absorb that, but I think most of all of us have been around long enough to know that's part of history and not strange. The snapback or not in 2023.
You know, one thing I might wanna just rewind a bit and preface my remarks on is that, you know, our tester market and our tester business demand is very sensitive to the growth rate of our customers. For example, if a customer of ours is growing units at 4% a year and creates a certain tester demand for us, if their growth rate nominally increases from 4%- 5%, small impact on their business. For Teradyne, that can be a 20%-25% acceleration of demand for us. Because our demand's driven off the first derivative, let's say, of customer growth. Obviously we're very sensitive to what the macroeconomic conditions are going to be in 2023 around the growth rate of our customers.
The technological issue around our largest customer in the smartphone market and the move to 3 nm, I think that's kind of baked in and is gonna happen. Macroeconomically, what's gonna be going on in the world and the inventory digestion or not that may be lagging from Q4 into early Q1 is hard to tell right now. You know, we're kind of reading tea leaves based on what we've seen happen, which I've said is, you know, maybe about $100 million of demand dropout to what we're anticipating will happen. I really wouldn't wanna get too far out on a limb, Mehdi, for forecasting 2023, but, you know, those are the factors at play.
Thank you. Thanks for the detail. One quick follow-up for Sanjay. Given all the color and guidance, it seems like this year revenues could be down mid-teens ±, and you're adjusting your OpEx growth of only 4%-6%. Looking to next year, if we get a snapback, you do benefit from an easy compare from 2022 into 2023. Should I assume that your OpEx growth would accelerate again, or should I assume that the OpEx growth will be minimal to, like, let's say, mid single-digit %, 5%?
Yeah. Mehdi, this year we had an anticipated plan, and obviously, as Mark said, we manage the business on a trend line, and we believe the fundamentals that we laid out in our January call and the earnings model are still intact in test and IA. With that backdrop, you know, we did have a plan to increase OpEx 11%-13%. As I noted, we curtailed that. We believe in delaying some expenditures that aren't gonna impact our long-term competitiveness. We haven't gone through the detailed planning in 2023. Of course, we'll provide that update in January.
My expectation is that we're gonna keep the focus on what's gonna help us be successful in the long term, and we're gonna for example, we're gonna keep on investing in industrial automation to secure the market going forward. We believe that market is still sub, you know, 5% penetrated. We have very good positions in our portfolio of companies that go to market in that fashion. You know, my expectation is we're gonna still operate at the kind of the 5%-15% there and still keep our foot on the gas to grow, to help drive that top line and capture more business. In the test portfolio, we're gonna continue to ensure we have a very strong roadmap and a good go-to-market.
The exact numbers I don't have right now, but the fundamentals of our business are the same. We're predicting to go through a little bit of a weaker second half than we anticipated, but you know, at some point, this will snap back which quarter that occurs. We're gonna keep our investment thesis intact.
And, and the-
Thank you.
Variable costs move with revenue as we talked earlier.
Right.
Yeah. Fine. Okay, folks, we are out of time. Mehdi, thank you for the question, and thanks everybody for their question and attention. If you have follow-ups, please reach out, and we'll look forward to talking to you in the days and weeks ahead. Bye-bye.
This concludes today's conference call. Thank you for participating. You may now disconnect.