Tech on Fire: Chris Grundemann, Analyst, GigaOm

Today we have the third in our series of podcasts for Tech Trailblazers called the #TechonFire interviews, in conjunction with GigaOm. This time Chris Grundemann gives an insightful interview all about networking, including how things such as digital transformation due to COVID has driven things.

Chief Trailblazer Rose Ross asks him all about the last 10 years of networking and how things have changed over that time. She also delves into what trends we’ll see in networking in the next 10 years, things such as the spread of AI and security will affect networking, and how automated tool will come to the rescue of network managers, not to take their jobs.

Plus, discover Chris’s worry about how youngsters coming to networking via cloud might miss out on learning about the underlying hardware, to their detriment. Watch the full podcast here:


Also available on

Interview transcript:

Rose Ross: Hello everybody, and welcome to the Tech Trailblazers podcast. We’re here with GigaOm and we’re doing a Tech on Fire series, which is very exciting, where we look forward into the future and look a little bit into the past as well. I’m delighted to be joined here from West Texas by Chris Grundemann. Hi Chris, hello.

Chris Grundemann: Hi, very well, it’s good to talk to you today. Rose.

Rose Ross: Likewise. So, you focus on networking; do you want to give us a bit of your background, because you’ve got an awful lot of experience. We’re delighted to be speaking to you about the whole networking side of innovation and how we’re seeing startups coming up in that space, and what we have been seeing and what we’re likely to see. But for anybody who perhaps has not had the pleasure of hearing you do one of your podcasts or spoken to you in the past, it would be great to find out a little bit more about Chris.

Chris Grundemann: Sure, absolutely. So, my background I guess is a network engineer by trade. I worked on some small ISPs in Colorado, wireless internet service providers was where I cut my teeth, and then worked on a network that’s now part of NTT and worked on another big network that’s now part of Lumen. I also then did some work in the cable industry, so at a non-profit called CableLabs, which holds the DOCSIS standards, and they do all the all the development work on behalf of the cable industry and did that for a little while. I was lucky enough to get my name on some patents while I was doing that and wrote a couple books along the way. And then I spent a couple years working with the Internet Society, and their goal is that the internet is for everyone. And so, I did a lot of traveling, I gave talks at Network Operator Group meetings and other types of events around the world, in 35 different countries. And then the last six, almost seven years, I guess now, has been more focused on Enterprise IT. So, bringing that networking background as well into the data centres and into the campus networks, but then also looking at security, and now more and more AI as well, and how that is going to play out in the IT world.

Rose Ross: Fantastic. Well, thanks for joining us, Chris. So, just to dwell a little bit more on what you’ve done, so, you’ve been a network engineer, and you’ve got eight patents or your co-….

Chris Grundemann: Yeah, that’s right.

Rose Ross: Co-named on a number of patents.

Chris Grundemann: Yeah.

Rose Ross: Also, you’re an author, and you’ve also been involved with an IETF RFC, I believe, as well.

Chris Grundemann: That’s correct, yes, I’ve done some work in the ITF, both from the angle of CableLabs, and then later on through the Internet Society, and so I do have one RFC out there with my name on it, which is cool.

Rose Ross: Yeah, got quite a badge of honour there. And you also podcast as well with a co-judge of ours Steven Foskett around Enterprise AI. So, tell us, network has come a long way in the last 10 years, what do you think have been some of the most significant innovations that you’ve seen in that, and obviously, IP has become prolific over the last 15-20 years. But what are you seeing as having been some of the milestones, I guess, in networking over the last decade?

Chris Grundemann: Yeah, I think you’re right, I think that just the proliferation of the internet and the Internet Protocol suite, TCP/IP, and how networking has gotten into every facet of life is definitely a big part of that. And we’ve seen, in large part, that’s happened in the last 10 years, maybe a little bit longer than that.

I like to say that the first 30 years of networking was figuring out if it would work, and then we spent 30 years making it work. And now we’re in a phase, it seems, that we’re really working on how it works. And what I mean is, it’s all about quality of experience, and ensuring that not just you have connectivity, but you have quality connectivity, right? Does this video show up clearly, can I hear you? Can the applications I’m using work, and can they work from wherever I want to work? And all of that’s obviously based on the network and the infrastructure underlying all of this IT, and again you’re talking of facets of all kinds of modern life now.

Rose Ross: Well, its certainly become a lot more important over the last 18 months or so, as a lot of us have been relying on the internet to just go about our daily jobs. What have you seen that’s been particularly insightful? You’ve talked about how ubiquitous it’s become throughout all aspects of our lives, but what what’s really allowed that to happen? What have been the big steps, the big leaps forward that we’ve seen in networking?

Chris Grundemann: It’s a good question. I think that one of the things is the basis is still a lot of hardware out there, right, switches and routers, which have grown up over time, and obviously higher and higher bandwidth and larger chips and A6 on those switches in order to allow higher bandwidth, lower latency, and all of those things. But, I think a lot of what’s happened in the recent last 10 years, and especially now going forward is the aspect of software in networking. And so not just the operating systems that are running on those routers and switches, but more and more, there’s software and programmes up around it.

One example is SD-LAN, which I think has been a big game changer I guess, over the last three or four years. Some folks started dabbling in it earlier than that and now it’s becoming fairly widespread, a lot of a lot of enterprises are looking into an SD-LAN type solution. And what that basically does is decoupling the control plane from that hardware plane, so that I can have a cloud-based network operations software that runs this network. And one of the interesting things there I think, is that a big part of what it’s doing is making the internet part of folks’ wide area network. And what I mean by that is in enterprise networking for a long time, if you rewind even further than 10 years ago, but back far enough, you really had servers in a closet maybe at your office, you had people working at desks in their office, everything was connected together with ethernet cables in that office, and you put a firewall on the internet connection, and you built private network to your data centre, then eventually moved those servers from that closet to the data centre. But there was still this area where it was your infrastructure. It was a network that you had gone out and bought these pieces of hardware and these circuits and put it all together, and you really had that ownership of this infrastructure.

And now with the advent of cloud computing, and virtualization taking over where we’ve moved a lot of applications, again, off those servers, we’re no longer necessarily putting them in a co-location facility, and then definitely not on premises at our own office. But we’re connecting to the cloud, which is of course infrastructure we don’t own, and we’re connecting to it over the internet which is infrastructure we don’t own. And so, understanding how to provide quality of experience over this infrastructure that you don’t actually have total control over, it is a big part of I think what’s going on right now. And that’s happening in the networking world, but also in the security world as well, where I just don’t have that perimeter where I can lock the doors and everything’s safe inside of that anymore, because there really aren’t any doors. People are working from wherever they might be, at home obviously lately, a lot more than they had in the past. And the applications are connecting to are again, out there on the cloud somewhere, they’re either SaaSi applications or applications running on infrastructure as a service, on one of these big public cloud providers.

Rose Ross: You touched upon security there as being one of those elements. And certainly, from I think previous discussions that I’ve had with some of your colleagues at GigaOm is around the fact that security perhaps was one of the issues that held enterprises back from the whole cloud engagement, really. So, something that wasn’t particularly important, you pop it out on the cloud, and everybody would be happy about that. Then people started saying, well, are we opening ourselves up to stuff? Can we putthere’s performance, theres security, there’s all of those elements. And so, what you’re saying is this whole software defined aspect has helped people to be able to not just deal with the performance, but also to help allay some of the fears that there may have been around the security about using the cloud, or do you think it’s more than that?

Chris Grundemann: I think that’s absolutely true. I think for sure that security was a big piece of this, and moving things to the cloud, and there’s still places and areas where you don’t want to move things to the cloud, for whatever reason. There are companies that are doing extremely proprietary things that they want to keep on premises, or they have some very specific requirements, and they need to do that. But in general, we have built the tools to allow most companies to be able to use the cloud in many cases. And definitely then, that is I think leading to… if you look at the trend of the SasSi, which a lot of people are talking about these days, where we basically added all the security features onto an SD-LAN platform, and now have this converged security and networking, which again that leads to cultural shifts as well as technological ones, because it used to be the security team, and the network team were different folks with different skill sets, and different budgets within a company a lot of times, and now that’s all been squished together.

And yeah, it’s interesting that I think the cloud is the reason that this is all happening, and also an enabling function of it, because the reason that the model of SaaSi works is because I can put these security appliances in the cloud and then of course they’re accessible from anywhere. So, people can be working from home or from a coffee shop, or from wherever they need to be, and still have all that full security. So, we’re really allowing this secure and mobile workforce to take shape. I think we were lucky enough that a lot of that technology was in place and at least available if not being totally used, 18-months ago when the COVID pandemic started, and has definitely taken off from there because it’s such an enabling technology for this, again, secure mobile workforce, I would call it.

Rose Ross: Yeah. Interesting you talk about the different teams and there being a cultural shift; which do you think has been the bigger challenge for companies? Has it been the cultural shift or the technology shift?

Chris Grundemann: Well, I think they’re really related. I think that the technology is definitely causing these cultural shifts that are required. And always, I think people tend to be the harder part of any change. There’s a lot of folks who learn to do something a certain way and want to continue doing it, there’s different lines of silos or whatever, teams and those boundary lines have been drawn. And so, figuring out how to work across those teams is super important.

One example is the virtualization that we did inside data centres with things like VMware or other technologies that then added network virtualization in there. And now you have this portal that used to be just where folks who were running the servers and storage would go to do things with their storage and compute virtualization. And now you add networking into that, and now you add security into that, and there becomes this question of, okay, well, who’s does what job now because the tools I used to use as a network engineer, were just maybe logging directly into the COI of these switches. And now I’ve got a virtualized network layer, whether it’s SD-LAN or something else inside of the compute layer, where it’s a different toolset and it’s a different mentality of how things work.

And so, definitely I think that’s a struggle and a challenge for a lot of folks. And not just within networking, and not even just in security, but broader. This whole shift to cloud definitely I think changed the way that a lot of IT operates, and we’re still dealing with that.

Rose Ross: Absolutely, I think a lot of people had to accelerate their digital transformation, they had to get cloud quick, its not about getting a cloud first, it’s cloud quick. Do you think that’s had an impact on some of the other aspects? Because moving speed doesn’t always give people the time to really ensure that things have everything in place. I wouldnt say it’s belt and braces, that’s probably a little bit extreme, because I’m sure isn’t that case, I hope.

But do you think that now is going to be a period ofnow that it’s been done, a lot of this has been done, are we’re going to have to look at retrofitting some of the other aspects that we’re like, Oh, maybe we should have done that’.

Chris Grundemann: Sure. I think so. It’s a constant evolution for sure, and I think that there’s still a lot of work to be done just even in just the general everything that’s folded under this idea of digital transformation and understanding that today almost every business is in some way a technology business. They may not directly be selling software, or have an app necessarily, but a lot do. And for lot, even if that’s not how you interact with customers directly, there’s still all this stuff on the back end that is how you operate internally, whether you’re using some kind of CRM, or some messaging app, or almost all of the email and calendaring, all those kinds of things have all gone online to cloud services.

But there’s still a lot of work happening in refactoring of applications and moving them there. And then again, connecting these traditional campus and datacentre networks up to the cloud networks. And I think there’s a big industry involved in doing that right now, so folks are building a lot more data centres with a lot more interconnectivity to be able to connect between data centres, to campuses, to cloud applications, and have that all work seamlessly. And so, we’ve got new players in the space that are building networks that weren’t ever something that was even imagined before, in order to enable some of this massive interconnection between companies, and between vendors and partners, and all those things.

Rose Ross: Well, youve given me a great segue there Chris. You talked about new players, and clearly the Tech Trailblazers is all about the startup community, the innovation that you see in that environment. So, do you think this is a good time for the networking and perhaps other startups who are maybe still quite fledgling, or have an idea now that may be is more pertinent because of the fact that we’ve got this much biggerwhat’s the best word to say? I want to say, big embracing of the cloud. Yeah, that’s probably the right word, is people have to give it a really big hug, and go, right, we’re in on this one now. Is this a good time to be a networking startup?

Chris Grundemann: I think so, yeah. I think again going back to that idea of software eating the world, and it’s definitely happening within networking as well, and because of that you don’t necessarily have to be manufacturing hardware in order to sell networking applications. That wasn’t always true, it used to be you went and got an appliance when you needed whatever it was, whether it was a router or a switch, or a firewall or a load balancer or something else, you went and bought some pieces of metal that were manufactured somewhere. And now a lot of that can be done with just software, and so that obviously opens the door for a lot more players to be able to come in and not have to spin up giant manufacturing processes.

I also think that, to our points we’ve been talking about through this conversation already, is so many things have changed, and it is a very dynamic space, which I guess obviously everyone’s always talked about IT being something that changes often, but I think we’re definitely seeing it on an unprecedented scale, and just the base architecture of how people connect things together is shifting. That’s all the way from – what we talked about already, one area SD-LAN, I think it’s an area where we saw some startups come in, and some of that has already consolidate a little bit. But now there’s again networking folks working with security folks, and security is becoming more important in all these things. There’s all the pieces and parts that go into that SaaSi model we’ve talked about. And whether that’s a CASB, or secure web gateway, or all these things.

Then in the security space, it blows up even more, and there’s I think, a ton of players coming into cybersecurity right now in all kinds of different avenues to try to protect all these different ways of connecting, and in different places that we have applications. And so, web allocation firewalls and load balancers and that stuff is definitely seeing a bit of a Renaissance, I think.

And then you move into where we get into the nuts and bolts of container networking, and microservices, and we’ve seen service mesh come out. I think almost everybody in service mesh space is a brand-new player to networking. They’re writing an interesting line between software applications, the DevOps type folks, and the traditional IT folks, and how to create this secure network inside of these container environments that are moving and changing all the time. So, there’s definitely a lot of different areas I think, where there’s a great chance for startups who are being successful, and who will be in the near future.

Rose Ross: Yeah. And because weve only had containers as a category for a number of years, because obviously doing this right at the beginning of when we started, but I think most of those have either gone on towell, I think the best most of them have acquired or are huge fundraisers.

So, we’ve seen a lot of success from our winners over the years, but they’ve been unusually consistently, well yeah, big raise, acquisition, dum, dum, dum. So, you’ve talked about that, and we’ve talked about where we’ve got to, I suppose right now. So, where are we going to see the next steps? We’re obviously going to continue to see a lot of innovation in the areas you’ve discussed. Which are the really hot ones? Which are the ones that you would be very interested to be seeing how they play out over the next few years?

Chris Grundemann: One thing we haven’t talked too much about is the infusion of artificial intelligence into these other systems, and I think we’re gonna see a lot more of that. I think that’s where a lot of innovation is going to happen, in both networking and the security space, is using the term AI Ops is now becoming fairly common, although the actual systems that are doing it aren’t quite all the way there yet, which I think is why there’s a lot of room for innovation still, is finding new ways to use these algorithms to assist us. And then really, it’s a co-pilot thing, I don’t think we’re going to see completely self-driving networks where you don’t need an operator at all, at least not for a very-very long time. But, I think we will see more and more areas where the systems we’re using are giving us helpful hints into what’s going on.

If there’s a performance problem that’s lurking behind some fuzzy data, or there’s some trend line that looks like it might do something bad in the next few days or few weeks. You’re going to get these more proactive alerts about what’s going on in the network, whether it’s just quality of experience or connectivity issues, or again the security piece. I think that is another piece that’s going to be really big, is that continuing convergence of networking and security. I think will play out in different ways, in different environments, in different industries, but I think we will definitely see and have seen security becoming more and more important in so many places. And that’s true of application developers, we’re talking about shifting security left. But, it’s also true in the network where just connecting the two wires together isn’t enough anymore. It’s really, we’re looking at things like network segmentation and encryption everywhere, and then how do we make sure that that all works out?

So, I think those are the two big areas; AI coming in and helping us, and security coming in and becoming more pervasive inside the network, are two of the biggest areas amongst some of the other ones we’ve already talked about.

Rose Ross: I couldn’t agree more. I’m very excited to see what we’re going to be having as entrants this year. So, weve talked about those elements, and I think that the automation, the AI piece of pretty much all of what we’re seeing, we’ve seen a huge surge in the number of AI startups who are coming and knocking on the door. And certainly, they do tend to fit with something else, so they may well have a security element to them, they may have a networking element to them. So, aside from those, one of the things I’ve obviously talked about previously with some of your colleagues, but IoT as well, obviously not necessarily your focus, but obviously that stuff that’s network, that people responsible for networks have to think about. What are the challenges from a networking perspective on that side of things?

Chris Grundemann: That is another big area. And the first thing is just the sheer number of devices being connected to the network is growing astronomically. And so, when you’re talking about devices that are meant for a human to interact with the system, there’s a limitation there by just the population. There’s only so many devices you can carry around; you might need a tablet, a phone, and a laptop, but that’s about the limit for most folks. But, when you start talking about putting sensors onto everything then obviously, the numbers are just staggering. So, just a general ‘being able to connect things’ is a big challenge there.

And then we’ve seen these new low-power WAN protocols come out, LORA is one, there’s some others, there’s some more proprietary ones and some open ones. But enabling these sensors to be able to send small amounts of data at really low bandwidth to keep the power envelope really-really small, so that the sensors can live… the goal is to have these watch battery type devices out there in the field for up to 10 years, because you don’t want to have to be going out and replacing these things all the time. And you can’t have them plugged into real power all the time, if they’re monitoring the temperature of a package in a frozen truck, I need to be able to move around, and move from truck to truck, and so it’s got to be running on battery power. So, being able to do that is super-important, as well.

So, between just the number of devices, and the types of environments they’re in, I think those are definitely networking challenges that are being addressed right now. And that’s a really interesting new space, we’re digging into it pretty deeply at GigaOm, and it’s emerging. It’s happening right now, and it’s one of the most fun things I get to do is try to tease out what that’s going to look like, because I think it’s not quite clear yet, there’s a lot of players with overlap and people figuring out where their niche is.

Rose Ross: Yeah, that’s the challenge, isn’t it? And also, for the people who are going to buy it. Because obviously the people who tend to buy the research that you’re involved with, and youre right, are the people who try to make those purchasing decisions.

Chris Grundemann: That’s right.

Rose Ross: I bet neither of us are going, Oh, I’m glad I don’t have to do that job, because that’s sounds like a nightmare over the next decade!’

So, what do you think they’re looking for, what’s going to be key for them? Particularly let’s say from a startup’s perspective, what’s going to help shift the needle for them, because its always a big step, isn’t it? Enterprise technology, when you’re responsible for that, there’s an awful lot of things with price tags that you already have to pay for, the basics. So, despite the fact we’re talking about how great it is to have innovation, and its all wonderful, and its a great opportunity for startups, but the end of the day, someone’s gonna buy this stuff. You can get as much VC funding as you like, but at the end of the day somebody has to have a need for it. Somebody has to have a problem that they’re willing to part with their hard-earned budget, or the company’s hard-earned budget, and there will be a lot of pressure in those purses right now because people are trying to save money.

So how do you stop being seen as a cost? How do you take what your new, shiny networking innovation is, and turn that into something that somebody at the enterprise level is going to be wanting to buy, or feeling that they have to buy? What’s going to be the silver bullet for that?

Chris Grundemann: That’s a really good question. My most basic advice there is a lot like the stock tip, which is to buy low and sell high. And what I mean is, the key is it needs to not be a cost centre. So, you need to be finding a way to save that person money, make them money, or lower their risk. And really, I think you’ve got to do one of those three things for a business, to be able to gain customers. Obviously, that’s very reductionist and pretty simple, and quite straightforward, but I do think that that’s going to be the key, is looking at what is the actual bottom line business change that I can enable with this technology, because your point right now is not an environment, and it really is an environment where I can just throw money at toys, right. And so, I really need to talk to folks about that business change.

And so, security is definitely something that’s changed – the dialogue around security has changed drastically in the last 10 years, and I think i will continue to over the next 10 years. And so that’s a thin area whereby mitigating that risk, you’re saving a company potentially a lot of money. And so those are ways to get in there.

On the networking side, again, it’s what is this transformation that this business is going through? And how can I support that, and again, either help them make money, or help them save money, and whether that’s higher efficiency, because I’m putting sensors on all the devices in a factory, and so now I can tell when they’re about to break down. That’s a really good reason to spend some money in that factory and get higher efficiency levels and get ‘mean time to repair’ down lower and things like that. So, I think it’s really about tying it to the business.

Rose Ross: When people pitch their technology to you, because obviously people want you to know what’s happening. How do you think we’re faring in the startup world, when you speak to technologists who have a new innovation or perhaps an innovation which isn’t 100% new but it’s a different take on something that’s already out there, are they’re getting it right, are they able to express themselves in that sort of way that can resonate to the enterprise IT buyer?

Chris Grundemann: Yeah, definitely some do. I don’t think it’s true across the board. I think that there is some innovation that’s happening today that’s almost innovation for innovation sake, and are solutions looking for problems. But I do think that for the most part, most folks are coming at this from an area of, they see an actual need and are trying to fill it. So, in general I think most people are getting it right. Of course, it needs to be fine-tuned over time, because those stories are going to change as the environment around it changes. So, I think generally though, most folks are tapping into these key trends and doing a fairly good job of it and connecting that back to the actual business need.

Rose Ross: I think that’s also helps that a lot of them have to pitch to the VC community. So, they have to keep things fairly high level and focus on those business needs, because you can’t guarantee that somebody who is going to give you money is going to fully understand the pain point. So, you need to be able to express those; what is the pain of your customer, and how do you alleviate that pain?

Chris Grundemann: That’s very true.

Rose Ross: We talked a little bit about AI, is that a bit of a silver bullet for this type of thing? Because AI feels like it helps save money. Now, everybody’s always gets a little bit nervous about that because it sounds like we’re going to lose headcountis that really true? A lot of people I speak to in the enterprise IT environment are saying it’s not really the case, it’s about doing the more mundane stuff that somebody would just be beating their head against the wall if they had to do it day in day out. Or being able to do more stuff that just isn’t humanly possible because you’re dealing with an attack surface, a threat landscape, for example in security, thats just so vast that you couldn’t cover, they just pick up the anomalies, and then somebody sifts through that, who can give the human eye and the experience to that, Oh, that doesn’t look quite right, that’s popped up. Would you say that’s correct?

Chris Grundemann: I do, I do think that’s right. I think that automation in general, and artificial intelligence, and down to the machine learning and neural networks as the deepest areas of automation, is a big thing. And I don’t think we’re talking about anybody really losing their jobs over it. And what I mean is, there’s so much work out there to be done now, as we’ve been talking about. All these trends we’ve been talking about, the internet and networking becoming more and more of every business, and digital transformation and security becoming more important. But the number of connected devices is growing, the number of people using these things is growing, the importance of it to the business is growing. And at this point, really-really good automation working everywhere, is just helping us stay afloat.

So, I think it does potentially save costs in that you maybe don’t have to hire a ton more people. But, there’s already more work than we have skilled people for, and so giving them the best tools possible, to your point, to take some of the day-to-day mundane tasks off their plate and let them do the bigger picture thinking, is absolutely helpful and is needed. And I don’t think we should be too worried about it as far as job security in IT.

Rose Ross: Yeah, there’s always been discussion about the skills gap. And we don’t seem to ever quite be able to make it any smaller, in fact I think were possibly making it larger, because obviously diversity of what we can do now from a tech perspective just keeps growing and keeps growing. Do you think that actual automation will buy us the time to get that skills gap addressed? Because it will allow us to perhaps encourage people to take on some of these more exciting roles, and more creative roles and perhaps more rewarding roles?

Chris Grundemann: I think it will, I think it has to. I think that we need this level of automation to stay as caught up as we are now and not fall further behind to your point. I think that we’ll do the best we can with it as an industry. I don’t know that we’re gonna see like the skills gap go away. I think that for a long time, definitely if we’re talking about the last 10 years or the next 10 years, I think for that for the next 10 years we will still see networking security job postings go unfilled for long periods of time because there’s just not enough people to do the amount of work that’s needed right now. And so, yes, these tools are going to be the thing that helped us get through that.

Rose Ross: And I always think the problem is that the skills keep moving forward, and we can never quite keep up. We’re chasing behind the curve of what’s needed, versus what is actually humanly possible to keep getting people to learn.

Chris Grundemann: Exactly.

Rose Ross: So, we’ve talked about the networking side of things, a lot of the good stuff that we’re seeing; where do you think we’re not seeing innovation that we should be seeing? Are there any gaps? We’ve talked about the skills gap, have we got a technology gap as well?

Chris Grundemann: That’s a really good question. I don’t know that I can think of an area where I see a real big need that’s not being filled at the moment. I think one area that doesn’t get maybe enough attention, not from a from a startup producing a product point of view, but just from a mindshare of the industry point of view is, the dirty little secret about software is that it all runs on hardware. And I think that we’ve definitely seen a shift in what most people understand as what IT infrastructure is. And so there’s a lot of folks who are maybe coming into IT jobs now, they’re maybe in their twenties and just out of college, and their job as an IT infrastructure engineer is to work on AWS, which is great and is needed, or Azure or GCP, or Oracle or whatever, but they’re working on some public cloud, where they’re basically just interfacing with a software interface that can use all this virtualization to give them the infrastructure they need. But they have no exposure to the servers or storage arrays or routers, or switches or firewalls, or the actual hardware that lives underneath that. And so, we may be making this a new and worse skills gap where that base level infrastructure, the actual connecting the switches together, connecting the routers together and knowing how that stuff works is something that I am concerned that we could fade away from that too much, and that there’s going to be less people available to do that kind of work than there has been in the past, because we’re looking at infrastructure of a different layer now.

Rose Ross: Yes. Well, I think parents will understand that when you try to explain what a landline is. What is that? You don’t need it, youve got a mobile, why would you need something that’s plugged into the wall?

Chris Grundemann: Right.

Rose Ross: Yeah, I think that’s an interesting aspect. And certainly, something I think we need to bear in mind is theres something about actually knowing that there are flashing lights, there are cables coming out the back of a box, and theres stuff happening physically somewhere, that is powering what you’re interfacing into.

There was one thing that we havent talked about, which I was a bit surprised we hadn’t. So, I’m just going to put that out there, we’ve already put out IoT stuff, and it kind of relates I think to a degree; the edge talking about edge networks, Core versus edge, and how we’ve developed these sub-sections within the networking world. Do you think they’re just going to be where its at? Not in a YouTube sort of way, but just generally speaking?

Chris Grundemann: So, I think it depends on what you mean, by the… Generally, yes’ish, but that’s because the edge is wherever it is, and maybe that’s silly. But there are lots of different edges, I think, and so there’s one way to look at the edge, which is pushing applications as close to users as possible to reduce latency. And that’s been the kind of job of what a lot of content distribution networks did in the past just for content, right. So, if you’re streaming a movie or something, it’s being streamed as close to you as possible. And now I’ve seen those companies are definitely adding new features and functionality to run more developed applications at the edge, and not just streaming content. So that’s one area, and then there’s another piece where that’s definitely happening in the mobile edge, where we’re pushing this out close of the cell sites themselves. And so that from that perspective, that edge is really about lower latency, and so better performance for users. That is again, part of what’s enabling some of these other technologies we’ve been talking about, the cloud security things that are happening – those happen at the data centre or PoP closest to you, to lower your latency. And they have to happen in a lot of areas so that you can move around and have a distributed company.

And then it does tie in with the IoT stuff as well. So, for example, again, if we are talking about the sensors in the factory, I may very well want to run a mini data centre in that factory, to be able to grab the raw data from all the sensors and cameras and things, right there live – run some artificial intelligence algorithms against it to understand what’s going on. And not necessarily transport that up to a public cloud or a big data centre somewhere else. And I want to do that locally, and then only store the things that I need to store somewhere else, to do processing somewhere else if I have to.

So, I think we’ll see a lot of different development in the edge, whether that’s again, the edge data centres that are where the CDNs are working, or the mobile edge where the mobile operators are doing this. Or like I said, I think some companies will be going backwards in that history we talked about, where they’re going to be putting servers back in the closet, so that they can do uncompressed research on the data.

Rose Ross: Yes, but that’s to do with bandwidth and speed, and efficiency again, isn’t it?

Chris Grundemann: Yeah, it is.

Rose Ross: So, we’ve looked back over the networking and touched on, well, quite extensively on security and AI and looked at that in the future. Is there anything else that youre excited to be seeing what will happen over the next few years, the next decade?

Chris Grundemann: Yeah, I think, the summary there for me from an operator perspective, from a practitioner perspective, is just the enhanced levels of visibility and control we’re getting. And so, whether that’s just from advanced telemetry, or we’re actually putting AI into the system, or whatever we’re doing, there’s more and more opportunity to understand what’s going on at a really granular level and have control over it. And so that I think is, across the board, a great thing that I’m seeing that will continue to advance and should move the industry forward.

Rose Ross: Brilliant. Fantastic. Well, thank you so much, Chris. It’s been really insightful to share half an hour or so discussing these aspects. I think it’s going to be a really interesting time, so I really appreciate you sharing some of your insights. I’m sure we’ll get to read about more of them over the next 18 months because I understand you guys are going to be very busy sharing that through reports. So, thank you.

That’s Chris Grundemann. He is with GigaOm, and he’s an analyst specialising in networking, but very much involved in security and AI, as well.

My name is Rose Ross, and I’m with the Tech Trailblazers, and this has been a Tech on Fire podcast with GigaOm.

Thank you, Chris. I really appreciate your time.

Chris Grundemann: Thank you, this was fun.

Rose Ross: All right then, speak to you soon, take care now.

You can catch more from GigaOm on our platform, which is or follow us on Twitter @techtrailblaze or find us on LinkedIn.