Arista’s Latest and Greatest: Innovations in Campus Since the Last Mobility Field Day

Name is Kumar. KI am the vp GM for the campus business for Arista. So I spend all my time worrying about LAN and wireless in particular and how we're gonna grow that business.

And I think as, as he mentioned, Arista is very well known typically for our data center and our cloud and an AI architectures and all the momentum, right? The buzz is always there, but I always say while the sizzle continues there, we are building a lot of stake in the campus side and the enterprise. So the company's two priorities today are AI and campus.

And you can imagine why we spend all this energy and really the priority to make campus, uh, very relevant, right? And mobility field day is particularly very important. We have many field days, as you know.

Uh, you will see us continue to invest in mobility field day, every day, every year over the year. So that's always gonna be a fixture, whether we do the other ones or not, because that's a very important area. We understand it's very critical for enterprise.

And I, you know, having spent a lot of time in land switching over the years with a history in Cisco, and there was a time when wireless was a tail and the tip of the spear was wired, but now it's inverted, right? Wireless really is leading the way. So that's the real, that's the priority that we have and that's the reason we are here.

Okay, so couple of things, uh, in terms of the agenda, so I'll just kick it off and then we'll have a very interesting topic with Ken. You heard him last year about, uh, about quality and how we build our architecture and technology around EOS, but today he's gonna be focused more on wireless and this is a very interesting topic to be, to be discussed. Then we have a number of oohs and ahs moment with the demos, uh, around wifi seven, around UPSK, Agni.

There's a number of technologies we can hopefully get through the, uh, the, the one which brings the whole portfolio together. Okay? Alright, so couple of things.

Um, just stepping back from a land and a campus point of view, uh, arista's growth today in campus, uh, like I see mentioned, we do occasionally get, we didn't know you guys do wireless or we didn't know, you know, you're just doing the wide business, you, but that's not true anymore. Uh, this year we are growing literally greater than 30% in wireless year to year already. We are on a roll and our entire campus business is growing north of 50%.

So just to give you a sense, and we see waves and waves of customers coming out of the incumbents looking for an alternative. So the momentum is there. We tend to do much better as, just to give you a view on the enterprise side of the house.

So franchise deployments, large scale, distributed campus, large headquarters. So you see us more and more, I call it on the G 2000, the global 2000 and the Fortune 500 class accounts. That's kind of where we have most of our traction because of the way we run the sales playbook.

Uh, couple of things, uh, you probably heard from, uh, from Ken last year. Some of the key principles that we apply to our wired and Arista, I call it the X factor, is also applied to wireless. Number one, do more with less, right?

Don't reinvent wheels. Don't try to do the same technology twice over. Good example.

Our cloud vision management platform is the same management platform, whether it's on-prem or off-prem. Our new identity system called Agni Arista guardian for network identity is the same stack on-prem. Uh, oh, it started in the cloud and now we are announcing the on-prem version.

Same stack. Don't balkanize the architecture built one singular architecture that can take it across the spectrum. Lastly, uh, a lot of focus on operations.

I think you're gonna see a number of that in terms of security, right? Which is really the two, three pillars of how we are gonna bring this together. Now, uh, in terms of very specific to wireless, right?

You are familiar with the Arista story, right from the beginning, we always took an approach of when, when the company bought the asset mojo back in 20 18, 20 19. What made it different was we focused on the security angle and the uniqueness of the webs technology that they had. That was really the driving center of gravity to bring the technology in house.

And eventually that became the wireless portfolio. And here we are with a broad spectrum of capabilities. Now, one point I would make in that is that that particular webs technology with the, with the radio today we have this feature.

You see that dedicated tri band multifunction radio. It's something we've had in the, our wireless architecture from the beginning of time, right? For seven years now.

So it's a very mature stack and I know you're probably gonna see a bunch of demos around how we do radio assurance telemetry, all the other management auto band stuff is done on this radio and not come in the way of the client radios, right? So a very unique feature, it's across the whole portfolio and we've had it for years and we continue to take it forward. The only other couple of points I would make is, as you know, the aristas approach for wireless, just to recap, is to disaggregate the stack, take the management layer integrated into cloud vision, which are overall wireless singular platform for the company.

Take the control plane of the ap, leave it on the ap aps have become very powerful. They are very capable and there's a lot you can do on the aps including running a mini firewall application. QE the aps are, can be leveraged to the hill so that we scale with the number of aps and as the radio technology scales and we don't monkey with that, right?

Lastly, the data plane. We use standard industry standard VXLAN protocol to really terminate all the wireless traffic into the network, I call it, into the EOS network and just carry the traffic in the a OS segment, which is really how I call, we fuse the wireless and the wireless architecture to look like one integrated platform that scales with the network, right? So as Broadcom scales, a chip Scale at the tri and three tri and four multi terabit scale, we can just scale with the network and we want the same unique architecture to flow from the biggest campus to the smallest branch.

As long as you have a switch and you have an AP and you have a cloud vision, you are integrated in. So that's the idea. Just a no controller architecture.

A couple of other things which you'll see is app qe, very important element. We spend a lot of energy really capturing those signatures of the traffic that come into the AP and then match it in the cloud, uh, in the AI ml engine to be able to sort of classify the class of traffic, whether it's uh, uh, zoom call or a, or a uh, Google, uh, video call. How do we really make sure that we give that experience and really get ahead proactively of problems that come in the network, notably in collaboration, right?

I think we all know in wireless the the troublemaker can always be, uh, collaboration, uh, um, applications is where, uh, you can get into trouble, right? So I think there's a lot of energy spent to make sure that that technology comes to bear, right? And lastly, uh, obviously cloud managed and on-prem managed.

One other thing you will see in this slide, which is new from last year, is the EEO s software agents. So we are bringing more EEO s agent services like Dox identity, VT a, all those stacks natively onto the e, into the ap. The goal being eventually the AP looks like nothing but an extension of the EEO s network, right?

And we get all the benefits of large scale validation of all the stacks, the regression, all the stuff that can and the team have built so that we get the full scale testing and be able to really set it up a very large scale. Domains in, as you know, a lot of wireless, especially in universities, higher ed, large layer two domains still dominate a lot of deployments. And really we want to make sure that we bring more and more EOS functionality into the AP and take it in that direction so that the whole architecture is unified.