We'll go ahead and get started. I'm Rob Mason, the senior analyst at Baird, covering advanced industrial equipment, and welcome to the Keysight Technologies presentation. Keysight is a cornerstone piece of our test and measurement coverage. It's the largest global electronic test and measurement solutions provider, and we're very glad to have with us today, Neil Dougherty, who's the Executive Vice President and CFO, and Kailash Narayanan, who's the President of the Communications Segment business at Keysight, which is their largest business. I think we'll just go ahead and get started with Q&A here.
Perfect.
if you'd like. I think, maybe I'll just start, Neil. The discussion around state of demand is obviously very topical. Just start at the 30,000 foot level. You had orders that had tracked about $1.1 billion per quarter prior to the pandemic.
Mm-hmm.
As we went through the pandemic, those ramped up to about $1.5 billion per quarter. That included a lot of maybe early queuing.
That's right.
Just to get, customers to get in the queue. We had supply chain constraints. We're now trending about $1.3 billion per quarter, that seems to be, maybe a stable level that you're thinking about right now, that you commented on both, the first quarter, the second quarter, and perhaps the way we're thinking.
Looking at the back half.
-about-
Absolutely.
Yeah, the second half of the year. Does that $1.3 billion capture most of the true demand signals at this point, do you feel like?
Yeah, I mean, I think, first of all, just reiterating the point, right? We do feel like based on, you know, the observable inputs at this point in time, that things have stabilized at this level for us. I think, you know, over the course of the last, you know, 12 to 18 months with the supply chain disruptions, we clearly saw some and built what we've talked about as some abnormal backlog, probably in the tune of $400 million-$500 million over that previous, again, 12 to 18-month period of time. I think there's been some of it, some of this is probably a response to that, people having things that are scheduled in backlog and therefore not needing to place orders.
I think, you know, the macro environment, some of the broader industry-related, inventory and digestion that's happening, I think right now we're seeing, again, things kind of stabilize at this level, and I think it'll be a good base for us to, you know, to pick up and grow from, you know, once the macro piece of this starts to recover.
$1.3 billion in orders last quarter as well, the same as the first quarter. Can you peel that back just a little bit and see and let us know, you know, perhaps what you're seeing under the surface? Automotive was very strong, you commented on, communications, a little more volatility. Semiconductor.
Yeah
segments and from a in-demand standpoint right now, not a shipment standpoint?
Yeah. Why don't I start, and I'll let Kailash give some deeper comments. I think as you highlighted, we saw, you know, pretty good strength in our automotive business and our aerospace defense business, were areas of relative strength. I think our general electronics business was very stable, and I think maybe that surprised a little bit to the upside, you know, given the nature of that business. Then on the downside, I think we saw a relative weakness in semi, which a lot of people are seeing.
Then in commercial comms, where, you know, a lot of our largest customers in that space have been very forthcoming and public with some of their issues around, again, inventory issues, workforce restructurings themselves that I think.
Our customers continuing to pick up, particularly in the second quarter, we saw more engagements. Their strategic long-term programs are intact, and we've seen more of the reduction in the manufacturing space. We have some exposure to that in the commercial Comms business, but the R&D programs, the long-range strategic R&D programs, both on the wireless and the wireline side, are staying more or less intact. Customers are now thinking about the 2024 programs. Yeah.
Maybe we just stick there a moment, Kailash. You know, Keysight overall has had much more of an emphasis on penetrating R&D test to a greater degree than production test. How do you think about just on the production test side, that part of the business stabilizing? There's a lot of reorientation of manufacturing footprints. I would think that that probably has some positive impact on your business. Is that activity still ongoing? Is that accelerating, stabilizing? Maybe does that present a risk if we get out into 2024 and some of that activities were to pause?
You know, we have seen some relocation of the supply chain, you know, partly because of post-pandemic effects, but then also due to geopolitical considerations. We are seeing a little bit of a supply chain relocation and new facilities come up outside of China. I think, longer term, we do view that as a positive trend. This year, however, because of the inventory exposure that, you know, people have, it's more of a longer-term trend than something that'll kick up manufacturing spend in the short term.
How do you think about, Neil or Kailash, how do you I often get asked around the cycle? You know, people think in terms of a cycle, but, in the test and measurement space, it's, you know, perhaps a little bit different. There's been maybe a few periods in history, you go back to the optical phase or the global financial crisis, where, you know, there was some abrupt slowdown in the business. Often get asked, you know, how deep, you know, how long is a down cycle in test and measurement? When you think, I'm just curious to get your perspective around that. When you think about the diversity of applications that feed into test and measurement at the moment, you know, how do you think about a cycle?
I think it's a hard question to answer because of the diversity of our business and the markets in which we serve. There's really no one singular driver or sector on the wireless side. On the wireless side, it's this progression from 400 Gigabit Ethernet to 800 Gigabit Ethernet, even 1.6 Terabit Ethernet. You know, R&D work is happening. On our electronic industrial side, you see that in automotive, the push towards electronic vehicles and autonomous driving. Semi, the move to smaller process architectures, as well as the move to ensure supply by re-onshoring semiconductor production are big sector themes. General electronics, I think you're just looking at the broader themes of digitization across.
There's certainly, everybody's aware of the buzz around AI, ML. That's not necessarily new, but in terms of maybe the demands it's going to place on networks going forward, I'm just curious how you see that Keysight's opportunity to participate there and how you're planning to serve that market?
Yeah, it's a great question, and as you first of all, we're excited about this new use case, and we're excited about the opportunity. As you pointed out, it's not something entirely new. If you look at all of the wireline innovations that have been happening, whether it's a 400G network or an 800G network, they've all been focused on what's happening inside the data center and to address an AI/ML type of a use case. This particular use case now is reaching an inflection point. It may have been just focused on a few players over the last couple of years, but now it's reaching an inflection point.
The large hyperscalers are definitely taking off, and their ecosystem is going to take off as this use case proliferates. Right now, the customers involved in AI ML, they're largely focused on ensuring the performance of their training algorithms in these networks, these AI ML networks. That's pretty expensive, and it takes a lot of time, it takes a lot of money, and we have an opportunity to really provide emulation capabilities. How does the network behave, the different configurations? How does modeling and training time and cost, energy efficiency, things like that, how can you model those KPIs before you configure and deploy a network? This is an opportunity for us.
We're seeing some early success with the large hyperscaler companies, we expect this to move forward. We also have exposure to the AI ML use case through our traditional business. When you look at things like GPUs, CPUs, we have exposure to that. There's going to be more GPUs, there's going to be more CPUs. Our semi business is going to be benefited from that. There's going to be more activity from a chip R&D perspective, there's going to be more AI ML-focused chips. That means more R&D spend in a chipset development, functional, chipset test. Things like that bode well for us. We also see a more conversion of technology from electrical to optical.
Optical is a little bit more efficient. It's faster. At the chip level, you have integration of optical technologies, things like silicon photonics. We have exposure from that perspective. Within data centers, you've got east-west traffic, data traffic, moving between the servers, and everything from switches to routers and these high-speed digital interfaces to take advantage of this opportunity. We do see this as part of our SAM, but this part of the SAM is starting to inflate.
With, I think last quarter, you mentioned a little bit of softness on the 400G side, maybe. Just thinking about how that pause plays out, do we see that recover, or is it led out by 800 Gigabit Ethernet, or, you know, do we move on to terabit? How that plays out?
You know, 400G initial deployment occurred, started to occur, maybe 18 months iterations. We do see 400G manufacturing picking up, you know, once this macro subsides. We do see 400G manufacturing pick up next year, while 800G R&D is already started to pick up, right? It depends on different hyperscalers, different companies. They have different network architectures. Some could deploy AI ML type of a use case with a 400G network. Others may need an 800G or terabit network. It kind of depends on how their existing network architecture is laid out.
You could, you know, you could have two lanes of 200 to get you to 400, or you could have two lanes of 400 to get you to 800. It all depends on the internal architecture of these companies. We see both 400G and 800G and Terabit all driving and catering to the needs of the increased data traffic, which comes from use cases like AI ML. Short answer, we do see 400G manufacturing that will pick up again next year, while we continue to see more R&D investment in 800G and Terabit.
If we just step back, maybe high level, within the communication space, is it fair to say, wireline has maybe not experienced the same level of near-term softness that the wireless business has, or is it fairly balanced?
I would say, the decrease has been on both sides and primarily related to 100G and 400G deployments. We do have exposure to optical components and switch manufacturing and things like that have taken a step back because of the inventory exposure. Some of our top customers are scrutinizing their spend on some of their programs. I would say it's been a bit more balanced. We do see, again, use cases like ongoing revisions of 5G, non-terrestrial networks, new use cases. 5G is finally diversifying outside of the smartphone use case into these new industry verticals, so that on the wireline spend going forward.
Just around 5G then, does the deceleration that you're referencing, is that then been more on the physical layer side, less in the application layer or protocol layer? How should we think about the various buckets within the wireless 5G?
Yeah.
Cycle.
Again, it would be, while the deployments are continuing, it's a little bit slow. That's the piece that you could say is has slowed down a little bit. The new use cases are continuing to expand. Non-terrestrial networks, industrial IoT, RedCap, things like that. You've got application layer expansion going on in automotive and aerospace defense, so we see that aspect of 5G pick up. Standards-based evolution, 5G Advanced, incorporating AI, ML in 5G, these things are seeing R&D investment. The number of frequency bands continue to expand. It's now in the tens of thousands. Test cases associated with that, we see spend with that. O-RAN deployments have not taken a step back.
In fact, regional O-RAN deployments, and it's going to lead to more improvements in the core network, which will address use cases like AI, ML, and then 6G research is continuing. We see many of the long range R&D spend vectors continue, while the short term, impact has been largely in component manufacturing, et cetera, related to the smartphone use case.
As these new releases come down 17, we'll look out to 18. How's the demand materializing, I guess, vis-a-vis, you know, what the customer profile looks like, the products then solutions that you're providing into that demand, you know, how it differs or does it differ from what you supplied early on in the 5G phase? Is it more software orientation? Do you still need the hardware platforms to be deployed into those new customers? Just a little bit on what the, you know, the profile looks like as we go forward.
Yeah, it's a good question, and the answer depends. I mean, if you look at something like a non-terrestrial use case, which involves satellite connectivity, that's a whole new ecosystem. You have new customers coming on board, and there are new solutions, including hardware that's required. You could say that that's a new portfolio or an extension to our 5G platform with new customers. Something like RedCap, which caters to IoT use cases versus the smartphone. That could again attract a whole set of customers. Somebody doesn't have to make a smartphone, but they could make a module that's catering to some use case in a different industry vertical.
That could be new hardware and a new customer base. We see the evidence of that. Then you have ongoing improvements in smartphone use cases, the Dual SIM Dual Standby use case or frequency band extension, improvements to battery performance, things like that. In the standard, that would be more of a software upgrade to some of the existing customers. It, it kind of varies based on the use cases, but suffice to say that 5G, the promise of 5G is to cater to these multiple use cases, and over time, we do see the number of customers expanding, even during this first half, the top customers have slowed down a little bit.
You mentioned satellite in that remark. Space in general seems to be getting a little more, you know, commercial space or space in general, seems to be getting a little more mention. Neil, in some of your commentary and just on the quarterly calls, you know, can you just Is that moving up in terms of the level of importance in that aerospace defense segment?
Absolutely. I mean, we've been in this business, actually, in aerospace defense, but also in the commercial space application with the LEO satellites and so forth. What's interesting is, the government entities want to have the current legacy satellite networks interface properly and interoperate with the new commercial networks, including new technologies like 5G non-terrestrial networks. There's opportunities in user terminals, there's opportunities with ground stations as a trend. I mean, there's, there are industry reports, market research reports that says that this business, the satellite market in general, is growing double digits over the next 7, 10 years. We're excited about it.
You know, we have both a portfolio from a commercial standpoint as well as from an aerospace defense point of view, to participate meaningfully, have a strong position in it.
Indeed. Maybe shift real quick just to the automotive business.
Visible opportunity that's in front of us is on the EV side of things. While there is AV work that's happening, the big demand driver right now is the electrification of the drivetrain. We see that playing out not just in the big auto OEMs around the world, but through the tier one supplier base as well. I think we continue-- I'll let Kailash talk about the auto, the autonomous drive opportunity here in a minute. I think beyond that, we do continue-
The electronics and the wireless footprint are continuing to expand. I mean, things like satellite connectivity, we have OEMs contacting us for how to integrate satellite connectivity into all cars of the future. These are use cases, lots of sensors, Automotive Ethernet, because some of the legacy CAN bus technologies are getting replaced with Automotive Ethernet. Where things like private networks are gonna be used in the future to automate things in their factories, that's an opportunity for us to participate as well from an enterprise private network perspective, especially with automakers. Excited about what AV and some of these additional vectors can bring in, but right now it's all about EV growth.
Rather than just providing tools to these industries, but really looking at the fundamental problems that they're trying to solve, and how all of Keysight technology, Keysight's technologies can be used in concert to help solve those fundamental problems, I think is allowing us to outgrow the market.
Yeah, and we've kind of talked about this at our Investor Day. We're really looking at the company as an innovation enabler or as a workflow enabler for everything from simulation all the way to live deployment. As Neil pointed out, is to continue to invest in R&D and determine what it takes to win these next cycles.
I think, Kailash, a manifestation of that, in my view, it was mentioned at your Investor Day as well. You have 500 or 500 and more, Keysight employees that are effectively embedded at your customer sites. That's, how unique is that? What advantage do you gain from, you know, having employees, that deeply into the customer set?
Yeah.
differentiate it.
Engagement. Again, shifting left, having a lot of the R&D folks, these are not folks that are providing support or application level integration. These are folks from our R&D teams engaged with customers on co-innovation. That is unique. It's not something that, you know, our customers take lightly or we take lightly.
Create a value that we bring to our customers and look to make sure that Keysight, as a firm, is getting compensated for, you know, for that value delivery. I think, you know, across the portfolio, there's, you know, there are places where solutions are more differentiated than others. You know, we work very hard to understand that and to price our products accordingly. I think, as I've just talked about.
Is, you know, certainly as things are more challenging right now, maybe minimum level of volume growth that would need to occur to be able to achieve those over the time period?
I mean.
Target 5 to 7-
Yeah.
Over the long period.
Yeah, we targeted this mid-single digit, you know, kind of, levels of revenue growth over the cycle. Obviously, we're in a period of time right now where demand is under pressure and we're, you know, we're not hitting.
There is a breakout session, afterward in Astra A. I wanna thank Neil and Kailash.
Great.
For joining us.
Thanks, Rob.
Thank you.
Appreciate it.