Great, thank you. Thanks. Welcome everybody. I'm Amy Sutter, Deutsche Bank's TMT specialist, and I'm very excited to host this discussion with Arista Networks. Obviously a company that's at the center of cloud networking and increasingly AI infrastructure. Today I have Mark Foss sitting next to me, who's VP of Global Operations and Marketing, and Hardev Singh, General Manager of Cloud Titans and AI Platforms. Very popular topic. Maybe just to start, to get a little from both of you, maybe if you could both just talk about some of the top priorities you're most focused on today and where have you been spending most of your time? I know it's different.
Personally?
Yes, okay, personally.
Yeah, I wear a lot of hats at Arista, but I'm more focused on the enterprise side of things. For us, one of our biggest growth areas is the campus market, the enterprise campus market that's connecting to users. We're really focused on not only going after large enterprise customers, but one of the focuses that I'm working on right now is enabling channel partners to help us on go-to-market for going after kind of medium-sized campuses. That's a kind of a big initiative for us. That's me in a nutshell.
In my role, I work closely with our large customers, really understanding the requirements, their roadmaps, their challenges, and then partnering with them to align our roadmap on products and really being lockstep with them to define our future products. That's where I spend most of my time.
Yeah. I know we'll go into that more. On your last earnings call, you raised your revenue guidance by roughly $550 million. I think investors were very excited about that. How much of that revenue is attributable to these AI Cloud Titan customers? Maybe just help us understand your visibility there and the deployment cycles you're seeing.
Yeah. We didn't break out exactly what the $550 million was attributed to. I think we said that at least $50 million of that was kind of our VeloCloud acquisition that we just did. The other $500 million would be a combination of all of our business, be it from general purpose data center to AI to other use cases. We didn't exactly break that out. Momentum is pretty strong as we alluded to on our call.
Is visibility different with enterprise versus Cloud Titans? Do you think?
It can be. Oftentimes, you know, it depends on, you know, visibility is oftentimes directly correlated to lead times of the product. Generally, oftentimes the Cloud Titans will buy our higher-end products, which generally have longer lead times.
Yeah.
You get visibility as to how long your lead times are in there. It's more product-related than company-related. Because the Cloud Titans consume a lot of our high-end products, we're probably getting a little more visibility from them at this moment, but that can change.
Okay. Maybe just on the AI spend, you've been talking about front-end and back-end, and maybe there's like historically maybe been a one-to-one ratio, but it seems like the distinction is blurring somewhat. I know I talked, we did an IR-hosted call last week, and this is something that a lot of investors were talking about. If you could talk about that from your perspective, Hardev, a little bit, to help us understand, you know, when you're working with customers, the front-end versus back-end.
Yeah, when we look at back-end, it's very crisp and clear. It's the network that's connecting to the GPUs, right? That's why even the guidance Jayshree is giving is like we have a clear visibility on that portion of the network that's connecting to the GPUs. Front-end gets a bit fuzzy. The reason is, yes, you hear of ratios like one is to one, even one is to two, sometimes one is to three. The reason is because front-end, you have connectivity to the general purpose cloud, you have storage in there, you have connectivity to the WAN, you have some compute there to feed that data into the back-end for training or inferencing. Depending on how a vendor is counting that front-end, you could have some variability there and fuzzy math.
That's why you see that the ratio kind of, you know, having this wide band.
I think for 2026, I think we've said that we're probably only going to give one number. We're going to combine the two. I mean, on the front-end side, front-end, even our customers have said, we can't tell the difference because the same products that go into oftentimes supporting the AI cluster versus the general compute. Our customers say they don't even know the difference. If the customers don't know the difference, then we can't either. That is why we're just going to give kind of one guidance number for 2026.
Okay, great. Maybe just, you know, we've seen large increases in CapEx, especially from Meta. It's been a, you know, had a huge increase in CapEx. When you, like, how, when maybe talk a little bit about the relationship with some of your larger customers, you know, what kind of visibility you have into their plans. You did talk about like how tight that relationship is. Maybe you can clarify a little bit like how that relationship works in terms of their planning around their CapEx spend and what visibility you have and what part role you have in terms of helping, you know.
I mean, my view on the CapEx spend is that from when they announce the CapEx to when it kind of hits, you know, vendors like us, there's a lag, right? Most of the CapEx is probably going into securing power sites, construction, and there's a lag there. It could take anywhere from 12- 18 months before you bring in the network. Working closely with them, the priorities are really around next-gen, 200 gig service-based products, 1.60 products. The priority is around reducing power. What are the other technologies like LPO, NPO, co-packaged copper? That's where most of the engagement is with them.
From a closed partnership standpoint, right?
Yeah.
The visibility can be a year out kind of thing?
Yeah, can be because these technologies are really kind of cutting edge. Liquid cooling is another very important aspect of technology that we're working on, right? Liquid cooling is there in compute today, right? I think in the future, with these new data centers, could be 100% liquid cooled. At that point, the network needs to be liquid cooled as well. Liquid cooling, 200G service-based next-gen, 1.60 products, optical interconnect with these new LPO, NPO, CPC technologies is probably the top three I would list.
Great. Maybe let's talk a little bit about scale-out versus scale-up networks. Can you, for some investors that don't understand the difference between scale-out versus scale-up networking models, explain the difference? Then maybe, what is your opportunity across both?
Really straightforward. Scale-out is the network that connects, you know, hundreds of thousands of these GPU servers or racks, right? That's traditionally been InfiniBand, you know, when it started, heavily now Ethernet. Scale-up is the network that is connecting the GPUs within a rack or a server, right? The incumbent player is NVIDIA. We have a proprietary solution, which is NVLink and NVSwitch. As the ecosystem of these accelerators increase to, let's say, AMD or custom accelerators from the hyperscalers, as that share kind of widens, there is a new opportunity for scale-up networking for Ethernet vendors such as Arista. It's a new TAM that'll open up, but it's still slightly out from a horizon point of view.
Right. I mean, is, I think, you talk about Ethernet versus InfiniBand and the opportunity being as maybe customers are using custom ASICs or they're using an alternative to NVIDIA. Is that the main reason then why people will be adopting more Ethernet in scale-up?
Ethernet scales. Ethernet is open.
Right.
You can dual source Ethernet too.
The smartest customers don't want proprietary solutions. You know, now we hear the AI cluster sizes from hundreds of thousands of GPUs to even a million, right? You hear some of the Cloud Titans talk about clusters that could potentially go to a million. At that point, Ethernet provides the scale, the openness of the ecosystem, and then it just gets more and more complex from a software development perspective, the features around congestion management, load balancing, visibility. Customers really care about that. Ethernet really has been there and continues to solve all those.
Customers in general are just more familiar with Ethernet, right? They know it. InfiniBand, they oftentimes don't have any expertise on it. They don't want to have to learn and curve to do that. It's also beneficial to have, you know, Ethernet everywhere as well, you know, versus having to, you know, manage an island technology. People just prefer to have this simplicity.
It sounds like the heterogeneous, using different ASICs or alternative to NVIDIA, that's driving some adoption of Ethernet. There are also other things that are driving Ethernet. We're starting to really hit the knee in the curve, I guess.
Yeah, I think InfiniBand was originally, you know, NVIDIA really kind of bundled the InfiniBand oftentimes with their GPUs and that strategy worked early on. Customers realized that they have choice and Ethernet interconnect has just the same performance. Guess what? I'm familiar with it and I can manage it easier. I'm going to move to Ethernet. At the really large scale, there's also a scale limitation on InfiniBand as well. If you want to go up over 50,000, 100,000 GPUs, you're going to want Ethernet. That was also a consideration for the large cloud guys, especially who wanted to have aspirations to go above and beyond. They don't want a limitation. There's no limitation on Ethernet.
Yeah, I think the debate between Ethernet and InfiniBand is over. I think Ethernet has won there for scale-out. I think the next opportunity for Ethernet is to now go after scale-up.
Okay, great. Maybe one big question I think that I've heard many times before is just the competitive environment. The one thing you mentioned, Hardev, is your relationship with your large customers and how you're thinking, you know, working with them up from an engineering standpoint. Maybe if we could talk a little bit about white box then versus your traditional competition with Cisco or, and now HPE, you know, how that's been evolving.
I can start, Mark. I think the landscape is still the same, Amy. It's not changed. There are roles where, you know, that it's complex from a network architecture, from a product standpoint. You know, the large customers always want, you know, sometimes two, even three vendors, right, in a role. The competition has always existed, whether it's branded vendors or white box, will continue to be that way. We continue to, you know, participate there, show our value, and win our fair share.
Yeah, Arista's primarily our primary value proposition and why our company exists to begin with. Our main pillar for our growth has been our software. Arista is a software company and customers really value that. I mean, we're really the only vendor out there that has an operating system that was designed and deployed as a 21st century operating system. These other operating systems from other vendors were either 1980s or 1990s architectures. Customers see a lot of value there. As Hardev mentions, most of these large guys want a dual source. They'll take these other vendors as a second source. Oftentimes they want a dual source using the same chip. They'll use the same, they'll standardize on one Broadcom chip and then they'll use two vendors that use that same chip, but then they'll have different operating systems, of course, and different switches.
As Hardev mentioned, there's not a lot of change in those strategies over the last six months.
With the acquisition of Juniper by HPE, have you seen a lot of opportunity?
I think what we saw when they were in limbo and there was, you know, for that year that went on where HPE announced the intention to announce, Juniper wasn't actually finalized. There was a lot of uncertainty in the market there. Uncertainty is not really a good thing for customers. They're kind of thinking, you know, are these products going to be around for the long term or not? I think now that that's closed, that will definitely help them. For that one year period where it was in limbo, I think that created a lot of uncertainty in the market.
Right. Yeah. Okay. Maybe to talk a little bit about Broadcom and the relationship there, just on the Jericho 4 chips, how does that enhance the scale-out opportunity for Arista? What advantages does that give you versus competition?
Broadcom is a great partner. Whenever you look at a new generation of speeds, to have the full kind of network portfolio, you want to have a mix of different chips, right? Tomahawk and Jericho. In the current 800 gig generation, you can think of the Tomahawk 5, right? And the Jericho 3. Jericho 4 now complements the Tomahawk 6. Both those chips are based on 200 gig service. When customers are going to look to build their, not just AI, their classic data centers, you want to have the same service across. We will have the same kind of architecture and we build the product with the Jericho 4, which will suit the AI league, AI spine, interconnect, data center interconnect. It helps us complete that portfolio around that same speed.
Okay, with Tomahawk 6, are you seeing trials right now with that?
Very early. The chip has just started sampling. I think the earliest we could probably foresee products is probably second half of 2026.
Okay. Yeah, and then maybe just let's talk about co-packaged optics. Just like, what do you think about the advantages and disadvantages of that opportunity, and where does Arista stand on CPO?
I want to start by saying that CPO as a technology is not new. It has been there for at least two or three generations. The promise of CPO, which is the same, by the way, for LPO, is how do I bring the power and the cost down for my interconnect, right? With these AI clusters, obviously now that interconnect is very large, right? You're talking massive size clusters need thousands of these switches and then tens of thousands of these optics. That's the promise or that's the goal that these technologies are trying to solve. Where CPO has challenges is around serviceability and the whole operation side of things, right? Pluggable optics, you get different speeds, you know, a port fails, you replace it, you have different vendors to choose from. CPO, you know, your failure domain will go from one port to the whole switch, right?
Like 800 gig versus 51T worth of bandwidth, right? We continue to evaluate that technology. We still feel confident that LPO or newer technologies like near-packaged optics or co-packaged copper are technologies out there that will help the goal, which is to reduce power and cost. We'll continue to evaluate.
We're agnostic to what, you know, if the customer demands it, we'll build it and provide it. We're pretty much agnostic to the technology. Myself personally and a number of other executives at Arista Networks, we went through this exercise 25 years ago where it wasn't clear when gigabit Ethernet was coming out. It was like, is it co-packaged gigabit Ethernet or is it pluggable? We didn't know then what it was going to be. It turned out that for gigabit Ethernet, at least going on to 10, 4,100, pluggable just blew the doors off of co-packaged. We'll see if that trend holds true in this situation here. Generally, past can oftentimes be a pretty good indicator of the future. Again, we're agnostic. If customers want to buy co-packaged stuff, we'll provide it.
Okay. I guess if we start to, we're starting to see more and more CPO products come out. Do you think that we're, how long do you think it takes for customers to maybe look at adoption? Is this like a year?
Depends on the customer. I mean, generally they'll trial things for like three or six months oftentimes. It depends. Yeah, I think there are two aspects. One is the technology being ready in terms of product offering. Second, to see once it's deployed in some certain scale, to see how the TCO performs from operations and serviceability and failure. I think there's two step kind of phases there to kind of then, you know, come out and say, yes, this technology works. You would probably see, you know, the takeoff of large deployment.
When you start to see successful deployments, that's when Arista would say, maybe we need, this is what our customers are asking for. So we'll.
Yeah, we base everything on what customers want.
Okay.
I'd say at least a couple of years out.
Okay.
Yeah.
You talked about 25- 30 enterprise and Neocloud customers adopting AI on so far.
Yep.
Can we talk about the ramp of those customers? What are some of the main workloads, use cases that they're going after?
It's a wide variety. If I had to say a trend, a lot of these guys are providing kind of AI as a service model. They'd be like, you know, maybe a startup or someone who's just, you know, wanting to build out an AI cluster that they could provide, that they can actually charge for a service. That's one model I'm seeing. I don't know if we've probably seen others as well.
Yeah. GPU as a service, sometimes building capacity in certain regions for, you know, large customers, sovereign clouds, maybe government mandated through, you know, where you have, you know, a customer already for what you're going to build. A lot of activity there. The architecture is very similar to, you know, let's say the large customers or the Cloud Titans, just that the scale is smaller. We see a lot of inferencing use cases. We see diversity of GPUs. It's a fun space and it's exciting.
What are some of the challenges that you see? Is it maybe because they're, you know, not the Cloud Titans that are so experienced with large data centers, or perhaps funding? Just curious what some of the challenges that you might be seeing, and does it offer opportunities to really go in there and help from an engineering standpoint that might be differentiated?
One thing that stands out, and Hardev probably has a lot more information than I do, is the ability to get power. They have, you know, there was one large one that we talked about on our conference calls previously that they had an ability to get power and they ended up getting some delayed funding. That's a big issue. There's a lot of other.
Yeah, and these customers also care about the performance out of the network, right? I mean, these are large investments, and to get the best performance out of the GPU, they really want the best of breed network. I think they're looking at, you know, Arista to provide that capability, that differentiation, the software stack that we have, and the bunch of features we've developed around load balancing and telemetry. The viewers see us as, okay, these guys are participating in running one of the largest, some of the largest AI clusters, and we get a fair share on the table to participate in those opportunities.
Okay. If there's some places you'll see bundling and those opportunities, you know, we don't participate in, but otherwise we do.
Okay. I was going to ask, where are we in the upgrade cycle to higher speed ports, like hundreds or gigs?
Yeah, there's always, you know, many upgrade cycles going on. They never end. You're, you know, I guess starting from the lower speed, I mean, the transition to gigabit to 10 gig is almost done. There's very little gigabit going on anymore. There is a transition going on from 10 - 25 currently, but it's still, you know, 10 gig is still, you know, at least, at least four exercises of 25 gig, but that's happening. There's the 40 gig- 100 gig transition, which is complete. There's pretty much 40 gigs now, pretty much dead. There's no deployments going on for 40 gig. 200 gig is, there's primarily one customer in the world that's deploying a lot of 200 gig. There's not a lot of other customers beyond this one customer, which is deploying 200 gig. That's kind of a, kind of a one-off for 200 gig.
On the higher speeds, we're definitely seeing some 400 gig, I think is, is fairly close to 100 gig. There was a transition from 100 gig- 400 gig, and that's fairly close now. I think you're starting to see 100 gig peak a little bit. 400 gig is starting to, I think, you know, have more momentum there. The one that's going the fastest right now is 800 gig. 800 gig has only been out, you know, shipping for like a year. Pretty much all of the new AI, you know, if someone's bidding on a new AI project, it's always 800 gig. 800 gig is growing the fastest, but it's, you know, the newest and the smallest. Anyway, there's a lot of dynamics going on there.
The trend I see, again, with the large customers is that all this traffic from AI is having a knock-on positive effect on the classic, if you will, data center, right? It's forcing some of the speed upgrades to accelerate.
Great. Okay. I guess also to talk about how important your observability and zero trust security solutions are for customers in AI use cases. Yes.
Yeah, security is very important, but observability is key. I think once, going back to the same point, as these cluster sizes become large, having good telemetry, good visibility into where, you know, I'm having network issues, you know, with this size of clusters, things are going to fail.
Right. Meta put out a paper where they're like the biggest, you know, link flaps or, you know, ports coming down was due to optics failures, right? Where due to HBM, you know, memory corruption errors on the GPUs. Having visibility there and then to quickly, you know, get the ports back up really reduces the overall job completion time for them, which is a direct correlation to revenue for them. They really care about visibility, observability, and that's where with Ethernet and Arista, you know, we have a, you know, a great portfolio of features in our software stack to address those, you know, challenges.
Is it typically then where observability always comes from the networking aspect of the,
Yeah, networking plays a very key role. CloudVision, which is our management platform, provides that visibility for the customers, not only on the network, we go all the way down to the neck on the GPU. You get an end-to-end visibility of the AI cluster to the interconnect to the backbone in one dashboard, which is CloudVision.
Right. Okay, maybe let's talk a little bit about sovereign cloud and service providers. There was no sovereign cloud contribution expected in 2025. How do we see that market developing, and what do you think helps gain there?
Definitely sovereign cloud opportunities out there, and we're involved in those. I think if you compare the sovereign cloud spend to what like a cloud titan were to spend on AI, it kind of dwarfs it. I think it's not about the number of opportunities, more of what the scale is. Generally, I don't think there's a sovereign cloud that I've heard of that's planning on going over 100,000 GPUs, but there is a number of others that are doing that. I think it's not about the number, but it's more about the size. Let's go to some other.
I think you said no contribution from Arista from sovereign clouds in 2025. That's not true. In fact, if you see, when Jayshree spoke about our, you know, we started with five customers to four, you know, we still are very confident on our 750 number. A lot of that is kind of coming from these 25- 30. Other customers out of the top four that we talk about, and where sovereign clouds are part of that.
Maybe on service providers, which parts of your product portfolio are best aligned with their needs from what you're seeing today?
Service provider side, you know, it's kind of our fifth out of five verticals that we do. We generally do very well in the service provider space if they're building kind of a data center architecture to support their infrastructure, and we do very well there. They also use us for routing use cases, and oftentimes they'll use us for like an MSP service. That's where the VeloCloud stuff comes in. They'll oftentimes use our Wi-Fi sometimes for a service as well, but there's a variety of use cases.
Yeah, the classic service provider in my view is slightly on the, you know, on the decline from a TAM perspective. I think, yeah, you know, our portfolio of products with our 7800 routers, both fixed and chassis form factor, you know, play well into that space. We continue to compete in that market.
How's market share for, I mean, continue to gain share? Like, who's the primary share?
Do we break it down or talk about it?
It's broken down by region or by customer type.
Yeah, I think tariff uncertainty has kind of come and gone. Are you seeing any impact from a customer buying perspective? Is it still part of the conversation with customers?
It can be, you know, because it's a very fluid situation. Obviously, there's a lot of uncertainty, you know, six months ago. That's one of the reasons why we didn't change our full year guidance last or two quarters ago versus last quarter because we were uncertain what was going to happen in tariffs. Fortunately, we are covered under the USMCA exemptions and under the communications equipment exemptions for tariffs right now. Our tariff impact has been immaterial. We haven't had to make any changes there, which has been good. As we've seen, the tariff situation can change very quickly. We're definitely watching it closely. Customers will ask about, you know, do we see any risk there? Again, we just say what I just told you, it's very fluid and we'll continue to react when it happens.
How's the supply chain? I mean, we've come in and out of an inventory situation also. Is it pretty not having an impact in terms of delivering product to customers? Maybe from a cloud titan perspective, you might've talked about their power issue is like their bottleneck, but it doesn't seem like there's many supply chain issues.
Yeah, supply chain continues to get better, post COVID. I think one of the big changes is that, you know, the chip, the chip, that the chips that we use, those lead times have come down on that, but they haven't come down to pre-COVID levels yet. I think that's kind of one aspect of the supply chain, which is still an issue for us. You know, supply chain is never over. There's always some kind of a shortage or there's always some kind of a surplus of things. It's always something we work on a daily basis.
I think we are well positioned from a manufacturing supply chain in terms of diversity. We have contract manufacturing in Asia, Malaysia, Vietnam, and then Mexico. We use three or four different contract manufacturers, and there's fungibility there in terms of products. I would say we are well positioned to address any kind of foreseen increases in demand.
I want to check if there's any questions from the audience just in case. We do have a microphone. Okay. Maybe on the margin side, is there a big, just, how should we think about AI product margins? Is it more dependent on the customer or the product?
Definitely the customer. Generally, Arista does volume pricing. The largest customers who have the most spend get the best prices. Right now, AI is a pretty concentrated market. It's generally the largest customers that are buying a lot of AI. Yeah, it's definitely customer versus product.
Maybe you mentioned this. We'll talk a lot, talk about the campus opportunity.
Yeah, my favorite subject.
Yes, you'd raised your target.
Yeah.
Can you maybe talk about what's driving that increase there?
Yeah, for us, the $50 million increase was really because of our VeloCloud acquisition that completed July 1st. That extra $50 million was really because Velo really rounds out our branch solution. We have the PoE switches and the Wi-Fi, which you can go in the branch, but the WAN optimization for the branch was a piece that was really missing. Velo really fits in there and really smooths that out for us. The campus is just, I'm so excited about it because in 2025, some of the analysts were saying that data centers are a $30 billion TAM, but they're also saying that campus is also a $30 billion TAM. They're the same size of TAM. However, from a market share perspective, Arista is small, low single digits right now in terms of share of campus while we're kind of 30% share in data center.
We have a bigger runway, I guess, for growth, a multi-year runway for growth in terms of campus. It's an area which people don't talk about a lot, but it's an area of our biggest TAM. When people ask about it, I get really excited about it.
I guess in terms of market share opportunities, how do you see that playing out? You know, is it a go-to-market as well, or does VeloCloud also maybe accelerate the opportunity to gain share? Expand on that
Yeah. S o, VeloCloud definitely helps us. I would say that about half of the TAM of the campus market is large enterprise, meaning Fortune 500, Global 2000. You know, we focus on direct touch to the end customer for those opportunities, which we're really good at because the same guys that sell data center, which is great. On the small side of things, we don't really focus on that market, which is maybe 20% or 25% of the market. The other 25% of the market, which is medium business, you know, that's an area of focus for us as well. Our go-to-market is more on the channel side. One of my big jobs right now is to focus on enabling channel partners to help enable them to bring business to us, kind of on the medium business side of campus.
That's really big because there's potentially a million campuses on planet Earth. Your Salesforce can't scale to have that coverage. You need channel partners on the ground to help you. That's been a big focus of ours.
Where would you say on the relationships with the channel, like where are you kind of on a trajectory of like first inning to n inth inning for those relationships and seeing the benefit in market share?
For channel enablement, it's still, I think it's still fairly early. I think a big chunk of our business is the channel oftentimes plays a fulfillment role for large enterprise, but our focus is to get them to do the heavy lifting and enable them to go after the mid-market and be very independent. We're having some success there, but I think it's still early. You could tell by our share and number basis, there's a lot of room to go. That's a big opportunity for us.
Great. Thank you. I know we're getting kind of close to the end, but I know you have the analyst day coming up. Anything you want to leave us with to entice us to be sure we're.
There'll definitely be some technology updates, how we differentiate in the market, and then maybe some insights into next year.
Yeah, I think, you know, we'll probably talk about 2026. I think there's a lot of questions on Blue Box. I know for a fact that we're going to be talking a lot about Blue Box and go in depth on that. That'll be exciting.
Great. Thank you guys so much for participating.
Thank you. Thank you. Thanks for having us. Thank you.