Welcome to the Rosenblatt Age of AI tech conference, fireside chat with Arista Networks. I am Mike Genovese, the cloud and communications equipment analyst. Super happy to be joined by some of the management from one of the very best companies out there. We have the Chief Financial Officer, Chantelle Breithaupt, and Brendan Gibbs, Vice President of Product Line Management. Nice to see you both.
Yeah. Thank you.
Hi, Mike.
Thanks for having me.
Great. I am going to ask questions. We have 45 minutes, but I also want to let the audience know that on the audience screens, there is a widget in the upper right-hand corner where you can type in questions that will come to me, and I will ask the team here any questions that I get from the audience. Please feel free to send them in. Let us get started. We are going to talk about AI, obviously, and Etherlink, which is Arista's family of AI-optimized networking high-performance platforms with advanced software, has gone from only four or five customers in 2024 to more than 100 customers today. Can you break that down more for us by customer type, whether cloud titans, AI labs, Neo Clouds, sovereign enterprise?
Where is the incremental kind of AI fabric revenue coming from, going forward versus the first half of the year?
Yeah, sure. I can start off, and Brendan will tag team these as we go through to make it conversational. Let me just take a step back, how proud we are to say, "Hey, we had this Etherlink portfolio June 2024." I remember I was six months in, we are at our New York Stock Exchange event, and not only did we have a great kind of analyst conversation, but we had a great product innovation reveal. I think it is just indicative of how we feel Arista was almost born for this AI moment. First, we had the cloud, and now we have AI. Even take the Etherlink portfolio you referred to, Mike. What a great, robust set of products unmatched by any of our competitors. We give all the different customers you mentioned, segments, options for their AI journey at a minimum. We are super thrilled.
Now to get to how are we getting to this 12.6 40% revenue guide that we are very excited about. It is coming from all these sectors. The thing is that we are seeing, if you are talking about what is changing the second half versus the first half, it is more of the same, or even versus last year. It is just more of, more Neo Cloud, more Titan, more AI labs, more enterprise, more campus. Although campus is not quite AI yet, but just generally in the 12.6. Those first few customers, we were talking at that time of InfiniBand, and now we are only talking Ethernet. The fact we have been able to go from those large customers to a breadth of customers, I think is super exciting and showing that for a pure- play networking company, we can serve any of those use cases and continue to drive the innovation.
Going into next year, we'll talk more about scale out, scale across, and scale up, and Brendan can give you more of that through your questions. We're just super excited, whether it's training and inference, front-end and back-end, scale up, scale out, scale across. We're going to meet the customers where they need and really help them achieve their AI outcomes.
Great. You touched on it a little bit, but just kind of flesh out more how you would characterize demand right now. Is this a really true, I mean, it feels to me like it is, a genuine expansion of deployment and plans and needs for the network and the data centers. We also have this supply scarcity issue, and in the pandemic, I think maybe that created, if we look back on it, for certain companies, created pull forwards. Is any of that going on? How do you as the CFO look at your dashboard and kind of tell the difference between expanding demand and just pull forward because we're worried about supply?
Yeah. I think that mostly it's from the customer engagements, I think, Mike, because if you think about it, we mostly serve large enterprise and larger. That's who you're talking large enterprise, Neo Clouds, AI labs, titans. For the majority, we've been having ongoing conversations, and so we would get a sense of if they felt they were pulling it in because of supply. I don't think that's the environment. I do think sometimes people are pulling in early because they want to make sure that they come to us as a company to help them in AI environments. So they might pull forward their enterprise renewal because of those, they might pull forward their scale across because they just have to get to outcomes faster. We don't see pull forward at all, but if we did, it would be more from a demand expectation.
We don't see it on the supply scarcity. My dashboard is customer conversations, it's project evolution. It's in the sense of it's 52-week lead- time, so I don't think customers are going to lean in to 52 weeks unless they're pretty certain that it's something. I don't think it's very easy to pull forward with those kind of lead times. We're very clear, very excited by the demand environment. Even this year in the guide, if we can get a bit more supply, we could even go a little higher than the guide.
Great. Well, look, I've covered this stock for a long time, longer than just the past couple of years. But I've gone through my own journey with Arista and AI in the last couple of years because when we were just first modeling and talking about the back end, I was looking at Ethernet becoming like InfiniBand, and I thought, "Well, where is the value of the software here? Does that change the story?" I don't know if it's front end versus back end and inference versus modeling, or if it's actually there's a different explanation for this. But the value of the software portfolio and the features and the new stuff you've put out has persisted, and changed my mind, and made me more of a believer in the last year or so. Can you talk about use cases for software versus blue box?
It seems like they're both growing in the company, and there's lots of use cases for both, but I'm talking too much. I want to hear this from you.
Brendan, you want to start?
Yeah, absolutely. First of all, I'm mindful of what Chantelle said. We announced our Etherlink portfolio back in June of 2024. That was 800G only. That was scale out. It seems so long ago at this point, but the portfolio has significantly blossomed, and use cases have expanded since then. What we found is that EOS, our operating system software, has such attraction for our customers, primarily because of the most basic value proposition, which is speed with quality. AI is essentially an arms race. All of these customers are racing to get to market, racing to sign up customers, racing to innovate at the frontier of models and cloud expansion and GPU as a service, et cetera. So having that software foundation for the network from us really gives them peace of mind.
It gives them quicker time to deployment, time to first token, and they just stay up and active. That's the first level, highest level deployment value. What we've seen is that not only the quality, as I'm saying, but also the intelligence built into the software, the analytics that we embed to give customers insights what the heck is going on in their clusters, that pays off in dividends over time, so we can unlock the greatest performance, the greatest GPU utilization, and the greatest insights to what's happening in their clusters. As they've gone from the initial small pilots of scale-out clusters, and they've made them large, EOS has grown with them, and we've been able to unlock that speed with quality paradigm for them so they can drive the scale of their clusters apace.
As they've found themselves restricted by one physical data center, and they need to scale across to multiple data centers and AI centers, we've grown with them there too. There's incrementally even more value within the EOS operating system software for them for that, with routing intelligence, with wire speed encryption, with more insights across a wide area. The expansion of use cases has fit with expansion of our portfolio from 800G to 1.6T, and now with more and more intelligence along the way, EOS has remained relevant and even more so over time.
Yeah. I want to double-click there on scale across, because it is turning into such a big market, and obviously things like 800G, ZR pluggables and the optical line systems and now coming the multi-rail amplifiers, those are supplied by the optical companies.
Sure.
You kind of had this impression, well, these ZRs will just plug into available switching and routing for companies like Arista. But there aren't just ports lying around ready. They actually need to deploy, it seems like, many new switches to do this scale across. Am I thinking about that correctly?
Well, in some cases, yeah. If I think about a scale-out cluster, because that's where it's grown from going back in time, and it's not that long ago, you'd have these modular high-end spine platforms with tons of ports, and that's one of the hallmarks of differentiation of Arista, these high port radix modular platforms. In the old world for scale out, all the ports just went south to a leaf node. But as the data center gets writ large across multiple geographically dispersed data centers and scale across comes into play, now you just turn some of the ports sideways, so to speak, to go for data center interconnect. What you have is instead of a spine remaining local, your spine becomes geographically dispersed. In some cases, yeah, you're right, you need to deploy new platforms.
In some cases, it's taking advantage of the high port capacity to maybe plug some of the ports in with ZR and ZR+ coherent DWDM pluggables. The great thing about us is that we work with a broad ecosystem of pluggable providers. We offer the true differentiation for that scale across domain of the encryption-based hardware platforms with all the routing software intelligence and diversity of optical plugs, whether it be from any of the coherent providers, we can work with any and all of them.
Does EOS is important for that entire now network that's one domain but across multiple data centers?
It's important, and more so than ever. Absolutely. Because now you need to arbitrate whether a workload stays local or remote. If we're going remote, how do you properly preserve and prioritize the transmission so that you don't have any sort of congestion? It's difficult but easier to manage congestion and have no packet loss locally. But now if you're going across hundreds or thousands of kilometers, it's even more difficult. If any problem comes up, it's going to be seen immediately. So the routing intelligence and analytics becomes even more paramount, even more complicated. But we've risen to the challenge, and we've got scale across now a third of our business as part of the most recent guide.
Right. That gets to the point, which is this is creating incremental demand for switches. It sounds like you need to use up the existing ports or anything you have left, but then put in new switches.
Yeah. We're identifying it as a distinct use case, and for sure it is. But also another way to look at this is just gigantic clusters. What used to be smaller inside of one physical building is now gigantic, like I say, where the spine becomes distributed across many data centers. So we identify that as a scale-out cluster writ large across many sites. But yeah, it's a new use case. It's expansion. It's more ports, more systems, more software intelligence, and more value.
Yeah. When we first on the street started talking about this market, which wasn't very long ago—
Right
I think that the vendors that are trying to jump to the top of the mind that are really going to benefit here, on the optical side, Ciena, and on the switching and routing side, Arista. Now, we've also seen on the optical side that it's helping Nokia and it's helping Cisco as well. But on the routing and switching side, in terms of actually moving the numbers, it seems that it's really working for you guys. Maybe a little bit at Cisco, but I guess my question is, A, do you think this helps all of the routing and switching vendors, or is it somehow more particularly focused on you because of the three major hyperscalers who are ahead on scale-out are your customers, right? Two longtime customers, I think one newer customer.
Is this something that's going to help all switching and routing vendors, or does it somehow more share that go to you, and you get more than your share?
Yeah. No, it's definitely a scenario where not all vendors are created equal in this context, because before you can take a data center and an AI deployment and scale it across, you actually need to win first that back end scale-out cluster. Arista is one of the most successful vendors in deploying large scale multi-customer back end training clusters, period. When it comes time to take that cluster and expand it across multiple geographically dispersed sites with scale-across, that's absolutely within our bailiwick. It's more than just putting on a cap that says scale-across on the top and saying, "I've got a routing box, too. I've got a coherent pluggable. I'm a scale-across vendor." Because they didn't win the scale-out deployments to begin with. It's not just scale-across is a perimeter routing one box or one DWDM transport node. It's the whole cluster writ large.
There's very few vendors who have had those type of large scale back end scale-out clusters, and Arista is one of them, especially from an Ethernet perspective. We're absolutely a market leader. Yeah, it's not a all vendors are created equal scenario. This very much advantages us because of our success in scale-outs, and now our intelligence and capabilities for scale-across, it just becomes natural.
Great. I guess as long as we're talking about this scale-out, which is what you've always done. Scale-across is already moving the needle, and I just need more help understanding scale-up for switching and routing, right? We've done a lot of work on what does it mean for optical and the optical content that goes in and replaces the current electrical content in these racks. I don't really yet understand ESAN and what that opens up for you guys, and I know that it's not happening tomorrow. It's pretty soon, but not this year. Just help me understand not just the timing, but actually what it is, and when it does happen, how big of an opportunity it is.
I can take a stab, and then Shantanu, I'd love for you to jump in, too.
Yeah.
At a high level, there's multiple industry analysts, but I've seen it marked as high as like a $20 billion market just for Ethernet scale-up networking by 2030. That's an industry separate analyst report, but that's compelling as a market size that we want to go after and pursue. We're putting our money where our mouth is, so to speak. We're investing in pursuit of scale-up networking for Ethernet. We see there's a huge opportunity for the non-NVIDIA ecosystem of Ethernet-attached XPUs, where you need some sort of facility to create that kind of coherent memory on the back end between all of the different XPUs in a rack. You need some sort of high speed interconnect, and we think Ethernet switching is really well designed for that. I think that first of all, it's an attractive market. It's potentially a large and lucrative market.
We're helping lead the definition of that market to begin with, where we helped invent and popularize the whole concept of Ethernet scale-up networking. One of the software leaders and architects from our company is the co-chair of the ESAN Network working group as part of OCP. Not only do we have the industry standardization, not only do we have the Ethernet prowess for consistency of AI fabric from scale-up to scale-out to scale-across, but we're also bringing to bear a lot of the hardware engineering prowess that we've got from our modular chassis over the last couple of decades. We're bringing that to bear at this new level of complexity for scale-up networking within a particular rack. The opportunity is significant. We think we've got a lot of software and hardware differentiation and unique expertise to bring to bear.
Yeah. The only thing I would add is that's a great technology description, Brendan, so thank you. The thing that I would add is think about when we were talking two years ago, kind of the Ethernet InfiniBand. There's going to be some segmentation and scale up, we believe. There'll be proprietary NVLink, there'll be some kind of white box equivalent for the lowest needs, and then there'll be a value add Ethernet, which we hope is us as the branded vendor. We see that segmentation maintaining.
We got to the point of announcing to being scale up because we had almost like a crowdsourcing from our customers, "Hey, Arista, can you think about getting into this space as an Ethernet leader?" There's a bit of a pull and a bit of an innovation, let's get there sort of combination between us and some of the customer base, that's where ESAN was born, and I think that's a great thing. It will not be material in our 2027 financials because we do have to get the ESAN standards and do pilots and trials. But once it gets going in 2028, we hope it's a great new segment. When we update our TAM, we'll give a shot as to what we think that can be.
Okay, great. I got a question from the audience, and I'm just going to read it. The questions that are coming from the audience, are different than the questions that I would ask.
That's fine. Yeah.
Let me just read it. Historically, Arista's discussed the customers who prominently use white box will continue to pursue that strategy. Is this still the case? What could happen that would allow you to win meaningful business at these customers?
Yeah. Generally, the white box environment is not an environment we'd want to go after because it's kind of overpaying for technology for places where it's low use in the network, right, Brendan? So the white box use case remains the white box use case. Having said that, there's probably somewhere a little bit in the middle where the operating system that sits on top of that will be a difference as we go and kind of maybe into these heterogeneous environments. So maybe that's where it gets a little squishy, for lack of a better word, in the sense of where that will meet. But we're not looking to enter the white box business. That has its own use case in any of our customer designs. Brendan, anything you'd want to add?
Yeah. The only thing I would contribute is we should think about why some of the customers went for white box to begin with. There are some that went white box even before Arista was a company. They're going to persist with that strategy. But I think white box persisted for those companies because it was high volume and it was low entropy. The complexity didn't change overnight. It was a well-known commodity, and it wasn't a lot of technological change or innovation, so they could just stamp out a high volume and save themselves a couple pennies. As we think about the AI trade, that doesn't necessarily hold true, meaning that there's so much entropy of technological innovation. So there's an opportunity for some of these customers to think differently. That's what we're hoping.
We think that the speed with quality value proposition that I referenced for our branded EOS-based operating system and platforms has utility in such a world where innovation speed is off the chart. Where white box may not be as attractive because you need not just a white box hardware platform, but you need an open source software, you need all of the features added to that by somebody, and then you need to find quality, and you need to find time and speed, and maybe that doesn't all work together. So it's possible that in the future, as Chantelle said, we're not going after white box, but some of the value drivers for white box may no longer be as valid, and we would love for them to join our ecosystem.
Another question from the audience, which follows up on this as well as the last question or the end of the last question that you commented on, Chantelle. There's some chatter that the nature of scale-up domain lends itself a bit more to the white box model. Is this true, and how would you characterize opportunities that are attractive to you in the scale-up domain?
Yeah. I will start in the sense of, we do see that segmentation coming, in the sense of a white box equivalent scale-up, a proprietary NVLink that eventually transitions to Ethernet, just like InfiniBand has done, and then the value-add Ethernet. Then maybe, Brendan, why don't you take us through, what's that value-add Ethernet scenario?
Well, the bear case, so to speak, like Chantelle's noting, where you might want this white box would be where there's not that many features. The bull case for us would be where our customer values consistency of operations, where they've already got EOS deployed for their scale-out network. So having consistency stamped out the same way for scale-up gives them a very rapid deployment. It's already qualified. It's already proven to work. So when turning it up for kind of that new world scale up makes it even easier. Because we shouldn't pretend all of the people in a non-NVIDIA ecosystem going with new XPU, it's all new, stating the obvious. Adding yet another new thing to the mix of an open source operating system of white box adds extra risk and complexity.
So the opportunity for us in that segment who might want branded, is that we can give them that speed with quality to get deployment and just remove one variable from an otherwise net new and complex thing.
Yeah, for sure. If you think about all these customers to whoever asked the question, and Mike, all of these customers are racing to get the most compute to get the best economics they can, and part of that is utilization. So whether it's scale up, scale out, scale across, we feel open source Ethernet is going to get them the best utilization, the best outcome, the best economics, and I think that's the ebb and flow you're going to see as they work through their business models. So we're very excited to help them.
Great. Let's move on to some other questions. Just the 1.6, which I think is trialing right now with large customers. You said volume production in 2027. I think there's a lot of things that all have to come together to make a new platform like this work. Just any more on the timeline of 1.6, and then as we talk about NPO and CPO and potential for open standards NPO, CPO, how does that layer into the 1.6 product over time?
I would say from a timing perspective, as you noted, trials are going on now, volume for next year. From a technology perspective, it's relatively straightforward in that we can still, with this generation, have both air-cooled and liquid-cooled options. So there's no sort of blocker from a technology impediment perspective where everybody's got to go liquid, not everybody's ready. We have air-cooled systems. We've already announced them publicly. For those people who want liquid cooling, again, we've announced that as well. A lot of the liquid cooling requirement is going to be driven by the compute nodes. We're not going to be the tip of the spear for that. People will have deployed and wanted liquid cooling for the compute nodes. That will likely be deployed before any of the networking gets deployed to interconnect them. That's, again, likely not an impediment.
It fits into the overall build-out that Chantelle and Jayshree have been very circumspect about describing for a long time, which is Data center build out and powering and facilities build outs, they're all out of our control, but a lot of that will be done before we get there. I think 1.6T is here. The optics are here for that from a local connectivity perspective and 1.6T ZR for scale across probably next year. With regards to NPO and CPO, it's a technology set of options that we're looking at very closely. We actually announced earlier this year at OFC, we announced something we're calling open CPO. CPO, of course, is co-package optics. We also announced at the same time technology called XPO, which is our pluggable, high speed, liquid cooled 1.6T option. In that case, again, we're trying to innovate for the industry.
We're trying to give choice. If a customer wants CPO and they're just bent on that decision, then hallelujah. In that case, we would offer open CPO, which offers serviceability and multi-vendor options for co-package optics so that you're not locked into any one solution, and it's not a value proposition of throw the whole switch out if something fails. Serviceability we think is critical there. But XPO is a complement to that in the sense that it's fully serviceable, it's pluggable, it's liquid cooled, like I said, and it offers every sort of laser available from short range, plus LPO and ZR+. We believe in choice, and we think that with 1.6T, with open CPO and XPO, we're giving our customers tons of choice.
Great. Chantelle, can you just talk more about supply? I am going to leave the question very general and not give too much of a wind up here. Except to say that after the first quarter conference call, investors came away more concerned about that supply could hold back the growth of the company and then seemed to feel better about it after the second quarter. But how do you deal with the challenges of supply, and how worried should we be?
Yeah, I think that Jayshree and I like to focus on the demand and how the demand is morphing and growing and shifting, and we are super excited about the demand side. We think it is definitely a demand driven market this year, and it does not stop on December 31st. We see it going into next year. So we are super excited. No one likes to hear, "Hey, you have got a lot of demand, and it is being a little bit constrained." I understand. We deal with the information we have in a moment. One earnings call we talk about it, the next earnings call, we are seeing some development. We try to be quite transparent, even though not everyone likes to hear it. We need to be honest.
I think from a supply perspective, we are happy that we have done some great work and our vendors and our ecosystem have done great work to help us get to this 40% revenue guide, which we think is tremendous. We are very excited about that. If we get a little bit more, maybe we can do better than that. But the supply is not unconstrained going into next year. It is a little bit of whack-a-mole like you hear Jayshree and I talk about. Hey, the larger we get, the more power we have, the more conversations we are getting. It is like this really great snowball effect in the sense of what we can get done. We are super excited. I think that I do not expect it to be a topic all of next year, but we did have to be a bit realistic.
I think you heard other companies after us start to say, "Hey, yeah, we see a bit of that too." We will work through it. But the demand, we are excited, and we will find a way. It is a when, not an if in our expectation.
I've worked in. It's been almost 25 years since I've worked in the industry. I worked at Ciena before Wall Street. I do remember some things, and I've obviously been following these companies since then. I know that you win customers, and that does create stickiness and opportunity for a while. Particularly given how fast things are moving these days, right? In a supply constrained environment, customers want a good solution, and they want to buy as much as they can possibly get, but they're always planning on the next solution. When I think about your revenue growth rate, I know you're not going to give guidance for next year until you do, and then we'll start lower and then we'll move higher. From my perspective, yeah, 40% next year, no problem.
I start to worry about 2028. I don't think that's a done deal yet, right? I almost feel like 2027 is the demand is there. If the supply is there, we're going to be good. I feel like in 2028, 2029, we have to keep competing with the NVIDIAs and the Ciscos and the Celestica and the white boxes, and we have to keep winning, always winning new deals for 2028 and beyond. Is that a reasonable way to think about things and how confident are you in kind of the next round of wins?
I think that I'm confident in the sense that if there's a market, we're going to do well. I can't tell you, Mike and team, what the market's going to be in 2028 and 2029. Everything I see from the industry analyst has it out to 2030 looks great. Maybe that's a little optimistic. The thing to think about is, let's say there's always a demand for networking, and we have this additional really great layer on the cake that it's more complicated and it's for AI. If we think AI is going to be something in some shape or form for the next five years or in perpetuity, we're going to meet wherever it's at. We're going to meet the front end and the back end. We're going to meet the inference and the training. We're going to meet the scale up, scale out, scale across.
The thing I think is really unique about Arista being only focused on networking and nothing else is that we'll cover any of those networking needs. Maybe one or two or three years, it's these large scale across. Then, hey, there's five years of really great enterprise inference, and wouldn't they love to have one EOS to put all their API Agentic AI units into it and have one really great experience and be able to cover against things like MISO? I think, Mike, it's wherever AI takes us is where it will ebb and flow, but I think we have a really good shot in any of those AI scenarios. I'm not going to guide 2028, but if there's a great set of things going, we'll be ready. I think that's our responsibility and our excitement as a company.
We're not distracted with other things. We're trying to prioritize.
Anything to add, Brendan, just from a-
Yeah.
Competing now for opportunities down the road?
You're 1,000% correct. We don't sit on our laurels at all. This is something that internally we're paranoid about. One of the things Arista has been amazingly successful at in the past, but never stop innovating for the future is the pace of innovation. From an enterprise data center portfolio, we've always had the most broad portfolio. Same thing from a hyperscaler and cloud perspective. We've got fixed platforms, modular platforms, shallow buffer, deep buffer. We've got something for everybody in every architecture that a customer might want. We've shown an ability to innovate at speed across a broad swath of the portfolio, and that's, as Chantelle said, because we're a networking specialist. We've got the heft to be able to execute an investment in parallel across multiple dimensions and across multiple generations. We just announced two years ago, in June 2024, our 800G portfolio.
We've already announced our 1.6T portfolio. We're continuing to innovate, and as AI has grown in use cases from just scale-out to scale across to next scale up, we're investing for all of those as well. That is what we see gives us that opportunity to continue to compete for every customer. To be sure, we've got long-standing partnerships that hopefully advantage us for future opportunities, but we're always going to keep investing from an engineering perspective to make sure we've got the right products at the right time, and from an operations perspective, to make sure we've got the supply chain products to ship when a customer might want them. They work hand in glove, and we've got both parts of that engine firing in parallel.
Okay, great. I was going to ask a gross margin question, and I got one from the audience. But I'll ask the audience this question. How do you think about gross margins in 2027 with the hyperscale mix? Is that more than offset by scale across? Any other puts and takes investors should keep in mind for long term gross margins?
Yeah, I think the number one thing to keep in mind is mix. That is the gross margin driver, and we try to be conversational about it every quarter. Mix is the number one item, and mix, we have two quarters of visibility, so going into next year. What I would say is, until you hear differently, the 60%-64% range is the range to assume, right? Unless we have a different set of parameters that get us in a different direction, the 62%-64% range is what you should assume. The only other one-off that could be 20 basis points here or there would be something like tariff refunds, where you have these unusuals coming in and out, but those are 20, 30 basis points at the most. Generally, it's mix.
Because we serve this wide segment of customers and industries, we usually try to balance all those things out. If we have an over-rotation on something, we'll talk about it, be transparent. It's still great to win those deals and have those deals and be in the 60s handles is my thought. But 62%-64%, unless you hear otherwise, is probably the band to think about.
Great. Then, I guess on operating margin, could you just answer? If I ask the question about operating margin instead of gross margin, anything to add?
Well, I think, hey, you know that we're only one networking business unit as the whole company. So in the sense of, I'll always ask Andy, Ken, and Todd, what do you guys need from an organic investment perspective? So that's R&D, sales, and marketing. Hey, ask what do we need to scale the company? But when you're growing 40% a year, that's a lot of absolute dollars, right, to add in the sense of you don't necessarily scale your headcount. So there's leverage to be had there. I think we'll guide next year as we get there, but I think we're pretty happy with the guide this year being kind of in that 48% to 49%.
Maybe there's a year of some investment that makes that different, but we're kind of in this high leverage state at the moment, so we'll see where we get to for 2027.
Okay. Just over the last couple of quarters, I've been a bit confused by price increases that can go through because customers realize that memory and other things have gone up. What I don't, and it kind of changes company by company, was do price increases, can they actually help the gross margin? Do they just keep the gross margin kind of flat when input prices go up? How do I think about the impact of price increases?
Yeah, I think it depends on the company philosophy. We've held 62%-64% as our range since Analyst Day, right? We've had it pretty tight and not really changing that for a while. Since then, we've had kind of this price increase conversation on some of the components, to your point, Mike. Our philosophy was two things, two-pronged philosophy. First was to try to measure twice, cut once in the sense of let's think of all the things we know of today and do one price increase. Let's not do the backlog because they're already priced. Let's go forward. That's a philosophy. The other philosophy is to not try to make money off of our customers for a moment in time that's beyond our control and theirs, which is a supply chain moment.
You can gain accretion in your price with some confidence if it's a tech valuation that you're giving them. You're giving them something more that helps their TCO, et cetera. Our philosophy is to only pass the price increase along that we're experiencing to hold margin. If it was 65% before in the price increase, we're going to try to keep it at 65%. We're not going to say, "Hey, we have a price increase, so we're going to give you the price increase and plus something to make our margins accretive." That's just our philosophy. I think it depends on the company's philosophy at the end of the day.
Okay. How should we think about deferred revenues in the second half of the year? You really raised guidance for the second half of the year. Just help us think about what deferred might do in the second half.
Yeah. I'll always say deferred can come up or down on any quarter. I will keep saying that because it's going to happen at some point, and that's just how deferred moves. Generally, I don't guide deferred, but I can tell you I anticipate deferred to end at a higher balance in Q4 this year than it did Q4 last year. Generally on the year going up, is kind of what I can tell you. But I don't guide, and there will be ebbs and flows every quarter. Generally, you should see it as, it's complementary to the P&L growth, it's another demand signal at some point in time it's going to come to the P&L, right?
Obviously, these things are moving, scale across, scale up, even scale out, right? New customers, we should think that there is going to be new projects that go into deferred revenue all the time, right?
Every quarter. The great thing is that there is nothing there that I am scared has aged past expectations. If I was to ask myself some questions, I would be like, "Is there something sitting in there that has sat there too long? Is there some issue that you see?" It has been a very healthy, within a quarter of expectations, generally flushing through. I think that just shows that it is working from our innovation, getting to acceptance by the customer.
Absolutely. Just to also just remind me how to think about 10% customers and customer diversification. As you win more customers, people are following. Are you going to have new 10% customers, are you going to have new 10% customers? Over time, I do not know if that will be as important. How do you think about it?
Well, I think it depends on the investor I am speaking to. Some investors like customer concentration because it means big wins, and some customers. Excuse me, investors do not like high concentration. I will not please everyone. Generally, what do I think? I think that it is getting more complicated to become a 10% customer because we are growing 40% a year, right? The denominator is getting quite big at a high pace. I think it is going to become more complicated to clear the 10% goal. Never say never. We are excited. We have at least two, or you said maybe 3-4 will come in this year. It will get more difficult, and I think just as we gain share in other areas, the 10% customers, I do not think you are going to see six or seven. Do we float between 2-4?
Yeah, maybe that lasts for a little while.
Okay. Just talk to us about purchase commitments, your perspective on where you are with those and how we should think about it.
Yeah. I think a couple things. One is, we're pretty, I would say, prudent, if that's the right word, but we will also lean in with the financials to do things like this. So I'm okay to lean into purchase commitments because I'm very clear why and what is at the end of that and what it means. We have about two quarters visibility, so we're leaning into two quarters, and Brendan and I spend a lot of time, what do we lean into based on history and what we hear from the customers. So it's educated and informed between the four quarters. But the great thing is with our portfolio, it can serve many customer segments. So we're not over-rotating on purchase commitments that only serve one. If something happened to change, we could use that product across many other customers, especially with the innovation coming.
I feel pretty comfortable with that process and what it means.
Okay. There is a question from the audience. I will just read it. Quick question. Is there any scale-up in product deferred yet?
No, we do not even have a scale-up offering yet.
That is what I thought. Okay. You have a lot of cash. You have a lot of cash flow. How do you think about M&A versus buybacks? Do you have capability gaps that you would rather buy than build? What do you think?
Yeah, I think generally from a cap allocation. First I go to Ken, Todd, Dan, DJ, Shri, what do we need to do organically? Okay, we covered that, and that is pretty much in our run rate that you are seeing. Especially with interest rates, what they are doing, we are getting a lot of good return on marketable security, so we will continue to do a bit of that. Now we are down to share buybacks and M&A. M&A, we are always looking for things that make sense, but the two things we want to make sure is that it is compatible with all the innovation we do and does not break it and fits the culture. If we see things there, we do not see any major gaps, Mike, to your question, but we are always looking for great talent and these sorts of things. If something comes across, we definitely take a look.
We look at things, I would say, every month, and try to assess if it's something. We haven't seen anything outside of the VeloCloud acquisition that we did last year that's come up yet.
Nothing's changed in terms of more of a strategy to bring optics more in-house. You haven't spoken about that. That hasn't changed.
I think that nothing's changed. I think we've got a lot to work on with just doing what we plan on doing. Unless, Brendan, there's something you want to add, but-
No.
Nothing's changed there.
We actually invented XPO Optics, then we essentially open sourced it. There's an open MSA with over 100 companies. That's an example of how we innovated but didn't hoard it and didn't keep it to ourselves. We're not an optical company.
What's the status of XPO, and is that driving any incremental switch and writing revenues for you yet?
Definitely not yet. XPO is an amazing tech. We've seen a ton of customer interest in it, and like I just said, over 100 companies have signed on to build products and be part of the MSA. What XPO is doing for us now is driving a ton of discussions for the longer term horizon in product planning. XPO solves key problems for customers in terms of density as well as liquid cooling at that density. It's being built into new product concepts, new product discussions with customers, but it's not part of any sort of revenue or deferred or anything like that in the last.
Okay. Chantelle, maybe just comment because I haven't asked you, and we've got one minute left.
Sure.
On enterprise and campus, the success you're seeing there, the growth rate, the guidance for that part of the business.
Yeah. Just quickly, super excited about campus, 5% market share gain mode, similar to what we did, rinse, repeat on the data center. We're all systems go. We've won lots of greenfield new logos just on campus, which tells me that we don't need to be in the data center first, which is great validation and recognition of our product. $800 million last year, $1.25 billion the guide. You can do the math. What's that, 45%, 50% growth? Market's not growing that way. Seeing lots of green tailwinds from our competitors. One competitor has a lot of M&A confusion on the roadmap. The other one has too many operating systems. We feel really good, and this is a steady eddy. This is just a focus. It's higher volume, lower dollar, so it takes time, but we're excited.
Okay. Well, that flew by. It's really a pleasure to speak with you both. I hope the audience enjoyed it. I enjoyed it. I look forward to catching up again soon. Thanks to everybody for joining us.
Thank you.
Thanks, Mike. Thanks, everybody.