Jabil Inc. (JBL) Earnings Call Transcript & Summary

Good morning, everyone, and welcome to the fireside chat with Jabil. I have the pleasure of hosting Mike Dastoor, who is the Chief Executive Officer; and Matt Crowley, who's EVP of Intelligent Infrastructure for Jabil. Thank you both for coming to the conference, and thank you to the audience as well.

Mike, maybe I'll kick it off with you and Matt, feel free to jump in here. But A lot of investors I talk to what's been more sort of surprised to them as has been the significant repositioning over the last 5 to 7 years for Jabil. As you sit here today, you have 40% of your revenue exposure linked to AI, including some proprietary capabilities like optical while the interaction in your power products via Hanley. So when you now think about the next 5 to 7 years. What do you see as Jabil's role in AI infrastructure landscape in particular?

I think, first of all, thank you for having us. I think the whole Intelligent Infrastructure piece today is going really, really well, strong demand from everywhere. I think it's Kudos to the team, Matt and his team here who've actually built a profile, which is not product-based. It's not just a single silo, it's across multiple capabilities across multiple sort of the highlights that we've seen. We're in power management, we're in service racks. We're in liquid cooling. We are in servicing and maintenance as well. So there's a whole bunch of things that we're that were involved in. And all of these are coming to provision very nicely, where you make an entry through one particular silo and very soon the conversation goes across. I think One of the things that we've been trying to do is it's all about value offering, it's all about an expanded value that you provide to your customers. And if there is a chance for us to do an end-to-end solution across the data center. That becomes an extremely sort of condition for customers. So I think going forward, the next 5, 7 years, I don't see anything happening to this whole AI. I don't think there's a bubble. I think things continue to progress well for us. I think the only thing we have right now is some of the capacity constraints. So we're constantly working on those next 5, 7 years. We'll probably be the most exciting years that I've ever seen and Jabil is vast.

Yes. And I would just add, the proliferation of complexity in the hardware that we're seeing now around AI plays really well into our strategy where we have specifically built capabilities versus being focused on 1 product, 1 piece of IP. So as things get more complex, we feel like we're better positioned.

Okay. How do you then think about -- when you think about the growth over the next 5 to 7 years or medium term? How do you balance that with margins and improvement in the margin profile as well? Because I think you've seen a significant margin improvement over the last few years, which you've been rewarded from -- by investors as well. But how do you think about continuing that while pursuing the growth?

Sure. So I think if you look at margins, and you referred to 6, 7 years ago, our margin was when you were at 3%, 4% look like a ceiling when you had 4%, 5% look like a seal and so on and so forth. Since that time, when we were 3%, we've repositioned the company, and what do I mean by that? We're no longer a contract manufacturer. I don't like that it has very negative sort of connotations. I like describing Jabil as an engineering-led supply chain enabled manufacturing solutions company I think the engineering piece, the supply chain piece are as important, if not more so than the manufacturing piece today from an engineering perspective, if you look at the offering that we have, we almost have 9,000 engineers in the company. That's a big amount. Not many engineering companies have 9,000 employees, 9,000 engineers in their organization. A customer comes to us today with a concept, and we're helping the customer design the product. We're helping them design for manufacture in most of the time when customers come to us, that design can be manufactured. It can't be manufactured at the right cost or efficiency levels and at scale. So we help them through that entire process. So engineering is a big one, supply chain equally big, especially I think our supply chain team is really good during normal times. During constrained times, as we're seeing now, which is probably the new norm -- they're really, really good because of all the relationships they build out because of the systems and everything that they've focused on over the last few years. And then the manufacturing solutions, you look at robotics, you look at automation, you look at engineering, the quality that we provide, our test environment, et cetera. It's a higher value offering. So I don't see 6% as a ceiling. I don't see 7% of the ceiling. I'm -- the organization is going to keep pushing, and we're going to continue to go up that value offering chain. And that will allow us to continue with our margin accretion story.

Maybe let's deep dive into the segments and starting with Intelligent Infrastructure here. Hyperscale CapEx. We've seen all the companies, cloud companies raised their CapEx spend outlook. But how should we think about broadly how Jabil has leveraged to the increasing CapEx outlooks from these companies?

Yes. I think when you think about the strategy we've deployed with a focus on capabilities versus products, -- that cuts across architectures, it cuts across customers, it cost big cuts across models. And so clearly, increases in CapEx across that customer base is good. Great news. But the way we've positioned the business and our capability around supporting customers broadly in that area, we think puts us in a great position to capture more than our fair share.

Okay. Okay. So maybe talk about areas that you can incrementally address within those opportunities on AI infrastructure. You already do sort of compute, you have uptakes, maybe talk about whether you see the more sort of incremental opportunities with hyperscales in helping them scale their AI infrastructure.

Yes, I would say that the opportunities are everywhere, right? At this point, it's not about share, it's really about keeping up with the organic growth of the entire market. But -- if I think specifically where I feel like we're going to have a really good advantage, we demonstrated a 1.60 LRO part at OFC, which is about 11 kilowatts, dramatically lower than current 1.60 power profiles. So I think as that gets into call, we're excited about that. We also have developed a partnership model as networking moves away from the top of the rack and to a more system rack level architecture. We feel like we're in a really good position to partner with some of the customers that we have. We're never going to compete with them, but delivering the ability to scale out and scale across via partnerships, I think, is going to be a great business for us. And frankly, when you think about our inorganic growth strategy, it's really to Mike's point on how do we create more value that customers are willing to pay more for and thus have higher margins. So I think -- we see the Intel transceiver business starting to pay off. I think Micros is going to pay off significantly, which is where we bought a company that has differentiated liquid chip capability -- and so right now, it looks like that part can cool incrementally up to 4 or 5 kilowatts versus competitors, and then handle where we now have a services organization, where we can deploy 50 kVA gear not something you can pull somebody off the street in Virginia to do and then service it as an ongoing revenue recurring stream. So really nice margins there, and we're headed towards the segment being at line, if not accretive to the enterprise.

Got it. And most of the -- so just to summarize, most of the incremental opportunities you're highlighting are better margins than what you currently see in the corporate.

Yes, absolutely. We are going to be very disciplined about the business we take, and we're going to balance growth with expansion of margins, and that's our focus.

Okay. So maybe starting with one of those, which is liquid cooling, which is moving from just having a nice to have to now being mandatory or sort of required for all customers. Just walk us through the broad capabilities that you have to address those requirements, including Micros, particularly sort of how you think about Micro in the next few years, delivering revenue for you?

Yes. I mean we -- so we have very intentionally built engineering and architecture teams that can address any requirement across XPU, whether it's liquid cooled or airport. So I think it starts there. We've got a transceiver part that's actually immersion cooled. So we have capabilities in that space. We obviously have talked in the past about our factories on the East Coast, preparing them for liquid cooling, which heard the call in Q2. Clearly, we've done a decent job at. So -- then you bring in micros where we can have differentiated levels of cooling. And so as you get to Tomahawk 6, et cetera, where there is a higher level of power required and more heat to dissipate, we feel like we've got an advantage there. And then it kind of rolls right into our DCI business, where we have the capability to manufacture CD at scale per customer designs were not and power equipment. So across the spectrum from a liquid cooled perspective, we have capabilities that we can address pretty much any of our customers' problems with.

And just to add on that, I think if you look at liquid cooling, liquid cooling is a way of getting into the door as well. There's a high demand for this capability. We seem to have a differentiated sort of offering. We often go in through the liquid cooling door and soon the commonization move to serve the rags, moves to power management moves to other parts of a data center, and that is the strategy. That's why we actually acquired micro, it wasn't for the revenue stream that Microscope by itself. It's across the board. It's enablement of an entire data center.

Matt, going back to what you referenced the power capabilities towards the end of the last one, a lot of investor interest that we're seeing on that front in understanding power capabilities that you have. So maybe help us understand the capabilities that Handle brings the acquisition brings to you? And how should investors think about the growth opportunity with Henley?

Yes. So handling was a really good fit because their approach was also not 1 of specific product but rather engineering capability. So we've got a really nice complement to the engineering and architecture capability that we had already created in the space. It expanded a bit so we can actually now design and manufacture at scale, power products like PDUs. -- medium Vault switch gear as well as getting the ability to service and deploy the product. So -- do I think -- if I were to refer to what expectations around growth should be, I would say that, that business much to the way Mike described the entry point we'll probably start off between the $200 million to $500 million range, but we'll expand dramatically from there over the next 3 years.

And are you seeing -- any changes on the lead times of these power products in terms of how you -- what's typical lead time in terms of you addressing customer demand today? Is the supply chain getting more constrained? Is that leading to somewhat maybe a pricing opportunity eventually as well?

Yes, it will have challenges, but a lot of it is very customer-dependent and model dependent. So for example, with our hyperscale customer, we build their LDS gear. We build it and deploy it. but they have a procurement power. And so typically, we'll get more than our fair share of parts in that space. So we haven't seen it really impact the existing business. Does it threaten potentially future business could -- but we have a really good supply chain also to Mike's earlier point. And so we're very proactive in addressing any potential shortages we see coming.

Okay. Got it. So maybe now let's flip or discussing uptakes and your silicon photonics assets that you have are quite well placed because everyone wants silicon photonics at this point, in the networking -- can you highlight what your competitive moat is at this point? And how is it helping you in terms of landing and expanding with some of hyperscalers?

Yes. So I would tell you that obviously, the Intel transaction gave us capability, it gave us capacity and it gave us a quick entry into silicon photonics, so we'd be ready for co-packaged optics. I think that there's the potential for a relatively dramatic moat with our LRO part that's going to pull 11 kilowatts going to out 60. So that goes into different calls across the next 1 to 4 months. So we'll have to see how it comes out. The Quals can take anywhere between 2 to 6 months. But we feel like that could be a major differentiator in the space. And then we also created a pilot line in Ottawa Canada for advanced packaging in order to be prepared for things like process development on co-packaged optics where it's super complex. We can develop processes in Ottawa and then deploy those at scale in places like Penang as we get more business. And then obviously, we've had a continuing really nice business in the networking space. and being able to now deliver a co-packaged switch is going to put us in a great position.

Okay. So maybe talk about timing for co-packaged topic I think there's a lot of industry debate about what that actual timing looks like -- what are

Yes, Jensen, it's today. I think if you look at the economics of it, the price curve has not come down in line for gig. So my sense is that there -- as usual, in video will probably be leading clearly, Will the entire industry follow I think there has to be some level of standardization from customers in order for suppliers to go and create processes that are repeatable versus onetime events. And so until you get to that kind of an ecosystem economically, I think it's going to take a while. So it's somewhere in between today with 1 customer versus x number of months with 50 customers.

Okay. Okay. And maybe let's just go a bit further down that that you discussed, like the CPU switch, for example. But before we discuss CPO specifically on the switch side, you do have capabilities in switching that extent both across Ethernet and InfiniBand.

Correct. We build better.

So how are you thinking about growth drivers for those individually -- and then when -- how do you sort of take that forward into what are you getting in terms of visibility from the customer in relation to a CPO switch?.

Growth drivers are kind of everywhere. I would tell you that the entire market continues to expand. We obviously do see products mix. And very careful not to try and reveal customer-specific data, but I would tell you that the mix between an Ethernet and InfiniBand solution has been kind of back and forth, not 1 big pivot. So for us, it hasn't had a big change in the business. Overall, whether it's a CPO architecture or Ethernet or InfiniBand, again, our strategy around capability has us well positioned to deliver all of those SP1 Okay. Okay. Good. Maybe moving to overall concerns that we've been generally hearing from investors, -- how should we think about the market share for cable with your largest -- current largest customer within into infrastructure -- there are obviously a lot of concerns around market share moving around when it comes to compute in particular, the largest customer. What are you seeing in terms of position with the largest customer? And how confident are you about maintaining share.

Why do you -- why -- what's the thesis on share concern? More competitors coming into working with the largest customer on the compute side.