61. User-Centric Connectivity Has to Innovate – Tech Field Day Podcast

The world of networking seems to be innovating rapidly. Well, at least it is. If you're in the data center, thanks to things like cloud and ai, the pace of innovation is exceeding our ability to keep up.

And then we go back to the edge where our laptops and tablets are still stuck in a time zone from the mid two thousands. And this episode of the Tech Field Day podcast, user-centric connectivity has to innovate. Welcome to the Tech Field Day podcast, where each episode we bring together a group of IT experts to discuss a single topic in the enterprise IT space.

This podcast features a variety of perspectives from members of the Tech Field Day delegate community, and is also recorded in association with some of our tech Field day events. Tech Field Day is a part of the Futureum Group, and this podcast is also being published on our sister company's website at Techstrong tv. In this episode, as we head into Cisco Live, we're gonna be discussing some of the areas that need to be focusing on innovation.

But before we do that, let's introduce today's group of guests so that you can get to know them just a little bit better, starting with Sam. Hi, I'm Sam Clements. I'm a technical solutions architect at WWT.

And my day job focuses on, uh, solutions around wifi centric technology and mobility based solutions. Ed Whedon, uh, network engineer, uh, in the enterprise space. Um, currently free ranging at the moment.

And of course, I'm Tom Hollingsworth. I'm the event lead for wireless networking and security here at Tech Field Day. Let's jump into this episode's premise.

No doubt you have probably, uh, gotten an email or walked into a store and seen that numbers are getting super fast when it comes to accessing things all over the place, but it kind of feels like that no matter how fast things get, it all feels the same. Um, uh, in, according to the old Henry Ford quote, uh, people really just want a slightly faster horse that eats a little less hay, and I feel like that's what we are being delivered in the modern, at least edge connectivity centric area. And I think that when you compare that to the way that, uh, data centers and especially AI networking are being forced to innovate at a rapid pace, it, it feels like we're being left behind.

So the premise for this episode is that edge connectivity needs to innovate. Uh, now I I think it's kind of funny, you know, Sam and I have been at Mobility Field Day, uh, here recently, so we've gotten to see where a lot of the people are trying to push edge connectivity, but I also kind of feel like some of it really felt like table stakes, right? And, and I know that we're coming into Cisco Live, and that's usually like the area of where like companies like Cisco and they love to release new products during that event.

So let's, let's pull out the crystal ball for a second and say, if a company like Cisco really wanted to innovate in these spaces, what could they do? Uh, you know, just put on your, your thinking cap. What would you like to see, Sam?

Well, I, you know, it's funny because I, I look at, obviously Cisco being obviously the number one networking manufacturer in the world. And I look at, you know, this, this chase of speeds and feeds and, and, and frankly at the edge, it always comes down to how we can push more bits and bytes. But there, there seems to be a lack of innovation in how we do that, even from a, from an energy efficiency and from a noise efficiency perspective.

We're talking about, uh, internet circuits that are, uh, multiple gigabits per second. And in order to do these things, you need to have these big honking loud boxes sitting out on the edge. And, and I feel like there, that, that, it sounds silly to say, you know, I want a quiet small box that does all the things that the big loud box does.

But apparently that's really, really hard to do because they don't do it. I'd love to see some innovation on energy efficiency, some innovation on how we get speeds and feeds and services without blowing everybody's ears out all day long. Um, and I realize that's sort of petty, but you know, it, it, it, there's a need.

Well, it's not really, it's not petty, Sam and I, and, and I agree with you a hundred percent. And here's the reason why I think it's not petty. And when you look at one of the, the main themes that comes out of Cisco Live every year, it's sustainability, right?

Like, like we, we've read news stories over the last year of Microsoft literally leasing, uh, nuclear reactor capacity from the site, formerly known as Three Mile Island to run data centers. I saw this ridiculous idea last week about launching, uh, uh, data centers into orbit so that they can be cooled by outer space, which by the way, does not work that way. But the, but the thing is, like, it's, it's one thing to talk about.

We need to be more, uh, energy efficient and more sustainable in this massive, you know, warehouse full of data. But I also need to be a little bit more efficient here at my house, because if I have to drop in one of those huge honking, like, you know, 120 millimeter fan copper heat sink routers, just so I can run two gigabit per second ethernet into my house, like, I, I want the sustainability there too, because I'm the one that has to pay that power bill, No doubt. Well, and look at the power consumption of some of this gear that's coming in, not look at like APS and whatnot.

We're talking about 60 watts down to the access layer. In some instances you throw you, you know, that number adds up really, really quickly. You're looking at a hundred aps at 60 watts a piece.

Guess what? That's a, that's a power bill. Yeah.

And I think the other thing with, with the sustainability piece is, you know, having worked in the enterprise space for as long as I have, um, you know, doing a lot of office builds and whatnot, one of the key things that we always have to contend with is getting the right amount of space to host the equipment that we want or that we need for a particular site. Getting the right amount of cooling into that space, that's always a fun one to deal with. Um, and, uh, power, you know, I mean, cooling power and, and space, I mean, those things are all at a premium, especially now, now as, as offices function the way that they do function, right?

Um, I mean, it, it, it, you know, one thing, I remember you, particularly when I was doing deployments day in and day out, it was always a cha, it's always has been a challenge. And none, none of that has changed. So seeing a vendor come out with new gear, great.

How power efficient is it? How cool, how, you know, quiet is it? Um, I would actually challenge Cisco and say there are other vendors in the market space that are dropping devices that can support 10 gig, 10 gigabit plus, right?

Um, you know, at least, at least for the routing edge, you know, that are very small and power efficient. And are you saying that not everybody who needs a 10 gigabit circuit has access to a rack with a full air handler system and, and HEPA filters? Probably not.

Probably not. What's funny is you look across other, other technologies like ai, where, where they're just like, damn, the torpedoes, full steam ahead, power bring it on, like where, you know, kilowatts and generators and diesels and blah and y, right? Yeah.

But Sam, here's the thing that I think that most people don't understand about that I'm, I'm kind of glad you brought that up, is like ai, it is assumed that power is limitless, right? Mm-hmm. Like, if I can bring another circuit into the data center to run this stuff, I can, we know the limitations of power over ethernet because you and I remember a time when power over ethernet was maxed out at like 15 watts, 15 watts.

So it's like, if, if, if it needs to, like, we need to like disable one of the CPUs on this access point so that it can actually run on power over ethernet. If you want it to run at full speed, you've gotta go run a, an a, a electrical E drop to it. Yeah.

Right. To it. Yeah.

And then, you know, we, we've gradually gone from 15 to 30 to 60, and you know, we've even seen some of the specs for like the, the big, like 90 watt POE stuff, which I don't know that I wanna touch device that's running 90 watt p oe. Like, I'm afraid it's gonna look like back to the future where doc gets shocked. But like the, the thing to me is we are working within the budgets of what we can do.

So the innovation has to happen where we know that we are never gonna get more than, you know, 60 watts to an edge ap, uh, in the best of circumstances. That's assuming we have the newest equipment. Whereas the AI people are like, yes, what if we could actually just put a lightning rod at the top of the building and have the power just shunted directly into the, the servers?

So I, I guess maybe my question there is, is one of the reasons why innovation has been falling behind so much in these areas is because we are working within self-imposed constraints, right? Because the other thing too, and I'll go back to this, is we're also constrained with the devices that we're using. So, like, for example, you know, Sam is, is very familiar with the fact that, you know, it, it took a long time for an iPad and an iPhone to have more than one radio in it.

You know, we, we talk about some laptop devices that have three by three, uh, transmit and receive antennas, but even laptops, which are considered heavy devices now still have power budgets. Because I used to own an IBM laptop that had a 210 watt power supply and would run for 57 minutes on battery idling. Like, if you did anything with it, the battery just went dead.

And that was like, oh, well, it's a laptop in name only. So are we, are we seeing problems where even the edge devices don't have full capability? So, I mean, if I could jam a 400 gigabit per second fiber card in my laptop, you could better believe I would.

Well, you, you, I think it's an, it's a, it's an interesting conversation when you start to, when you start looking at it through the lens of things like cloud manageability, when you start looking at, okay, how, what, what is my management orchestrator on top of all of this network? And, and, and, you know, throw a Cisco hat on, right? Because it's Cisco Live, right?

Look at the, look at the, the behemoth that is Catalyst Center, uh, versus adopting a management platform such as the Cisco's cloud or the Meraki dashboard, right? There are legitimate power constraints to each of those, um, trajectories. And I feel like the, the boxes that are in play certainly impact that power efficiency depending on sort of which way you're gonna go, right?

You build out a rack full of d uh, catalyst center clusters, and guess what? That power bill is gonna be as expensive as your AI bill is. And, and that's not necessarily a fault of the big ass loud router that you're trying to manage.

It's because you've tacked a bunch of other stuff on top that requires something else to do the compute for you. I'm glad that you mentioned Catalyst Center, because some of us are old enough to remember Cisco works and, and the behemoth that that was, it's like, you know, my, my, my compute bill just quadrupled because this is gonna take 18 servers and two copies of MySQL. But I think that's actually one of the things that Meraki got right from the start was people don't wanna run the infrastructure.

They wanna access it, but they don't wanna maintain it. And so, you know, I, Sam, you mentioned when we were getting going here, like Di Meraki acquisition is 10 years old. Like, like, we are starting to see the fact that, you know, as we've heard over the last couple years that Cisco live, that that, especially when it comes to the wireless side of things, Meraki and Catalyst are one company now it, it is Cisco Wireless.

And, and that integration is super important because it creates consistency for people. Like I, I can remember trying to transition from using Arrow net aps to lightweight access points. And I, I made a, a year of my career out of doing that for customers and how big of a pain in the neck that was, and now it's consistent.

Like, we don't have these arguments of, you know, is it autonomous versus controlled? Is it this, is it that it's just wireless? Now we're still solving hard problems there, and we're still ensuring that users have maximum connectivity, but we're not arguing about which dashboards being used or what, what functionality is gonna be enabled.

It's just kind of all there. So maybe the question is are, do we need more things like that? Do we need more solutions that concentrate the functionality of management into something so simple to use that we can spend our time more wisely solving other problems?

Yeah. Look, you know, the Cisco's biggest competitor there for a decade has been Meraki. Um, and which is, which is, you know, obviously a, a, a tad on the alarming side.

Uh, and when we see Cisco, I, I think start moving, uh, cloud managed architecture back into the campus where they, obviously Cisco has a tremendous amount of core competency. And, um, there's some challenges there, right? There's some challenges with folks that don't wanna spread VLANs all over the place.

There's some challenges with tunneling technologies and the like, um, we've seen for many, many years that, that cloud manageability and, and campus, uh, haven't yet been fully fleshed out. And, and, and there's a need there, right? There's a need for an, an, an edge box or a data playing box or something along those lines.

Um, it'd be really great to be able to cloud manage my campus, and that is from top to bottom, including my, you know, my big ass chassis switches, my modular switches, you know, controllers, aps, fabrics, and the like. And today we're just not there yet. I think that's a big challenge.

And that's, uh, uh, I mean, it's a challenge that we faced where, where I, where I was, um, you know, and I think a lot of organizations do face, um, and I think, Sam, to your point that Cisco's biggest competitor is Meraki. I never thought of it from that perspective before, but that's a, that's a great way of putting it. Um, you know, Meraki did a lot of things very well from the cloud management space, and a lot of other vendors in in, in the world are, you know, starting to copy that or have been, have been doing it their own flavors of that out there.

And it's, it, it still feels like Cisco is kind of competing with itself a little bit. Um, not that Cisco would ever do anything like that. Uh, Sorry, that was my outside voice again.

Um, but it seriously though, I mean that's, I think that's where I think I almost feel like cloud management, uh, cloud management solution is almost table stakes now. Um, you know, especially any organization that's just, you know, spinning up right now, they want, you know, cloud management capability, um, you know, for whether it's for their products, you know, within their cloud presence and AWS Azure, whatever, uh, you know, but they want that same, that same capability, that same ease, um, you know, for managing their infrastructure, uh, you know, their on premises, uh, infrastructure. Uh, and Sam, to your point there, you know, there are pros and cons that come with that.

Let me throw a wrench into this conversation because you know, ed, as a longtime network engineer before I got this job, I completely agree with you. Like, I, I want cloud management, right? I wanna be able to log into a portal and manage things.

But other things that we're hearing in the industry are that people don't even want that. They just want it managed. And so when you look at companies like Nile and Brahman and Meter, they're going one step beyond what Meraki offered, which is you don't even have to manage it.

We will do that for you. You just Right. You just do things.

And, and I think that now that, you know, to Sam's point, you know, Meraki was the largest competitor of Cisco for the longest time. And, but now that, that they're one big happy family most for the most part, um, now the competition is going to be the companies that stepped out a little bit further on that edge and said, well, what if we just did it all for you? And, you know, we have heard from, uh, Nile and we have heard from Meter at various field day events, and I know that there's hesitation from the delegates about how to implement this stuff.

And boy, if you haven't seen those videos, you really should watch them. I, I, I, yeah, I'll definitely need to go back and, and, and watch those. 'cause that's, but I think that that's an interesting space.

But the question that I have is I, is is that because we're getting pressure from the C-suite to remove even more roadblocks in including things that we don't think are roadblocks? I, I would, I would take it a step further and say that it's not just pressure from the C-suite to remove roadblocks, remove roadblocks, move faster, less people, less spend. And whether that's, and, and, and, and, and whether that is a, a, a valid argument, it can easily be debated, but you can't deny the fact that there are pressures coming from top down to really cut costs, uh, across the board.

I, I, I'm speaking from direct experience with that. Um, you know, and that's, that's a big pressure point. Um, you know, and I think one of the big pressure points there is, is you get pressure getting pushed down to say, cut cost, you know, do more with less.

How do you do that? And the traditional ways of management that we have done management in the past don't scale. They just don't.

And so that's where I think you start moving in towards the cloud managed infrastructure and the MSP market, uh, whether it's a true MSP or like what you're talking about with, with Nile, you and, and other, you know, other companies in that space, what they bring to the table, you know? So it's, it's a very interesting time right now. Um, so yeah.

Um, I, you know, the, as a service thing is, is really comes down to a consumption problem, a at, at the end of the day in my book. But, you know, just because Uber came out doesn't mean that General Motors stopped selling cars to the, to end users, right? Um, you know, just because you have an, as a service offering, it doesn't necessarily mean that the world is gonna change overnight.

Uh, look at the, look at, look at what cloud did to data centers and look at the pendulum of swinging back the other way with repatriation of data and the like. Just because it's cloud and just because it's as a service does not necessarily mean that it's right for you and your company. And it doesn't mean that it may ever be right for you and your company, nor does it mean that it really doesn't mean anything.

It's just another way to consume the technology. And I don't know that I look at it as being a black and white. You either do this or you don't, or, or you do that, right?

Right. Solution for the right consumption model for the right use case. The age old, the age old answer.

It depends. It depends. I believe Sam has that trademarked at this point.

Wait, you, Sam, not, not Yvonne. No. He got there first.

The, uh, the, the, the joys of trademark registration, right? But, but I think that you, you all bring up really good valid points here that anyone who believes that there is a one size fits all solution yeah. Is going to be very, very mistaken.

And, and we're even seeing that, I, I dare I say it in the AI space, right? Like it used to be that everybody ran InfiniBand and mm-hmm. And the reason why is because InfiniBand was optimized for HPC and it also was owned by Nvidia, who was the large, is still is the largest manufacturer of AI stuff, but it was also a very one size fits all solution.

And I know that I really p**s people off when I say this, but it was basically the fiber channel of networking, right? Like, as long as your network looks like this, it's perfectly fine, but if you need to scale past this many nodes, well, you're kind of screwed. And so calling the token ring of networking, no, no, no, because I really, I really don't want to get the IBM people on my case.

But, but the, the people finally realized that they needed to expand past that idea. And so that's why we've seen innovation with ultra ethernet and Spectrum X and things like that, which are creating, uh, you know, they're busting through the scaling limitations, of course, they've got their other problems to solve. 'cause we all know that ethernet is probably the worst protocol for deterministic delivery.

But, but it's, it's that complacency, right? Like this is the way that every network is gonna look, and this is the way that things need to be built. I mean, anybody remember Cisco, uh, reference designs?

Um, you know, s was it srds like solution reference network designs that, like I, I still have them like wallpapered in my house. So you like, here's the core, here's the access, here's distribution, here's the internet block, here's the WAN block. Think I think I have some of those burned into my, burned into my brain at this point.

And now we don't do that. Now we have clove fabrics and now we have, uh, you know, very intelligent edges that do a lot more processing these two, it's almost like the colla the, the distribution layer has collapsed into, you know, something very unlike what we've seen. But it took innovation in the market to get us there because, you know, the fact that an access point is effectively like a very small server now, um, look, right?

The amount of, uh, processing that it has to do, the amount of policy enforcement that it has to be done at the edge because our edge clients are getting more and more massive. I mean, I had a conversation with somebody just the other day about running AI workloads at the edge. 'cause we now have the capability to do that.

Is it this where it, I always laugh because like you, if you look at the, the market for smartphones pre 2007, they all looked vastly different from each other, right? You had the Palm T and the Blackberry and all these other things. And then for some reason after 2007, they all looked the same, uh, little Donnas script, black rectangles.

Um, but it was because the industry was still trying to figure out what they wanted to do. And then when the winning idea came along at everybody immediately wanted to do the winning idea, are we seeing a situation where for the longest time, up until maybe a couple years ago, the winning idea was these traditional network designs. And then when a disruptive technology, whether it was cloud or AI, came along and blew that whole thing up, now everybody is scrambling to try to come up with a better solution.

And that's what we really need, is we need some kind of really disruptive innovation at the user edge to bring that along. Uh, use cases of course, evolve. And sometimes the technology is looking for a problem to solve.

I look at, um, edge, uh, containers on aps, right? Uh, that, that have been a thing you can do for the past, let's call it seven, eight years or so. And, and really what was the use case for that?

Well, hooking to some sort of external module or something, or doing crypto mining or something, right? The, the use cases for when containers came out on aps are, are essentially obviated today, right? They're, they're, they're, they're, they're, they're just, they're obsolete and they're not even being used yet.

That technology still exists. And so what are we now going to do with that? Well, now the buzzwords are ai.

And so now we're seeing, um, uh, p qc capabilities being put into, uh, boxes so that you can run a service that, you know, further protects your data. And everything is about, you know, this inline AI inspection for protection. Um, the, the, the use cases change and evolve.

And, and lord knows the same thing we were using containers for seven, eight years ago isn't what we're using them for today. And same thing on edge, uh, same thing on switches that have container support and routers and all of the devices, right? Yeah, I think there's some, uh, I think the innovation piece or, or that, that disruptive piece, I think is something that is starting to come to fruition depending on where you are.

Um, and I think that, I think some of that, I think some of that disruption needs to happen in order for that innovation to get pushed along. Um, the container piece is one that's very interesting, Sam, that, that you mentioned that. 'cause that's something that I was just beginning to explore in a previous role.

Um, you know, seeing what we could do with that. Uh, and you know, one of the big things that I always had in the back of my mind was site survivability. 'cause one, one of the big problems, one big challenge, uh, and particularly at an enterprise level when you're talking about, you know, a, a decent sized enterprise, um, is site survivability, at least for offices, right?

Like, how do you actually make sure that hey, I lose connectivity to my data center, okay, but now where, you know, all those other core services, how can I get those? You know? And so there's some interesting, uh, edge to that point of edge cases.

So I guess what we probably need to say, you know, 'cause this episode, if you're watching it when it releases, is releasing right before the keynote at Cisco Live, I want you both to pick one thing that you hope that Chuck Robbins and G two Patel are gonna release at the keynote that is going to wow you solve this problem. Make everybody happy. Um, Sam, I'll start with you.

One thing that you really wanna see, Uh, we haven't talked about it at all in this particular recording, but we, but the time is right for an outdoor wifi seven ap Yes. I I think that, uh, anyone who goes to any kind of sporting event, uh, will agree with you. Yes.

Uh, that, that we need something that can, uh, that can work outdoors and, and honestly, with all the craziness around six gigahertz, wifi, seven, all that other stuff, yes, the Time is right. Regulatory requirements. Yeah, there's all sorts.

And, and, and, and, and it's, it, it's, it certainly makes sense why we haven't had one just yet, but it's now we, we need it. Yeah. What, I mean, how else am I gonna be able to stream live stream my Beyonce concert experience without the right stuff?

Exactly. Ed, what about you? What's the one thing you wanna see?

Oh, I would think I, I actually, I think I would probably lean heavily on probably better, better cloud management without needing to do on-prem services. Like actually putting that control plane, uh, into cloud services. Kind of like when Mera, you know, from the Meraki model and have better unification across the entire platform and routing, switching wireless.

I think that's pretty fair. That's, that, that's kind of where I would, I would say that's table stakes now, right? Is, is unified management experience, uh, because they're getting killed with other companies that are doing that.

Alright, well I think we're gonna go ahead and wrap it up there. So, uh, thank you very much for listening to this episode of the Tech Field Day podcast. If you enjoyed this discussion, please leave a comment on our YouTube video, uh, subscribe in your favorite podcast application of choice, uh, so you don't miss an episode and consider giving us a rating and a review.

I'd like to take a moment for our guests to kind of, uh, give us some information about where you can follow them, uh, for their thoughts and content around these things. Starting with Sam. Uh, you can follow me on Blue Sky at Samuel dot Clement, Uh, also on Blue Sky at Avalon Hawk, uh, and also on LinkedIn at least gonna start posting a bit more there Standing.

net. You can also read my blog there and do all the other things. Uh, this tech podcast was brought to you by Tech Field Day, which is your home for IT experts from across enterprise it, part of the Futurum Group.

You. Let's check out our upcoming events as well as more great episodes at this podcast. com/podcast or view us on Techstrong TV or in the New Techstrong TV app.

Thanks for listening very much. We'll be back next week with another great episode. Until then, take care.