The EDGECELSIOR Show: Stories and Strategies for Scaling Edge Compute

Finding and Fixing the Next Bottleneck for Edge Computing with Dave McCarthy of IDC

October 10, 2023 Pete Bernard Season 1 Episode 7
The EDGECELSIOR Show: Stories and Strategies for Scaling Edge Compute
Finding and Fixing the Next Bottleneck for Edge Computing with Dave McCarthy of IDC
Show Notes Transcript Chapter Markers

Join us on an enlightening journey with our special guest Dave McCarthy, as we navigate the complex world of Edge and Cloud computing. From his early days as an IT practitioner to his current role as a research vice president at IDC, Dave has accumulated a wealth of experiences and perspective that’s sure to challenge your preconceptions. What were the stumbling blocks Dave faced during his career, how did he overcome them, and how did his progression shape his views on Edge and Cloud computing?

This episode is a deep dive into the challenges and opportunities of deploying solutions in the heterogeneous environment of the edge. We’ll tackle the complexities of hardware, communication protocols, and data formats, and explore why the need for scalability is pushing companies up the stack. And if you’re wondering about the sustainability of cloud computing or how to power small, battery-driven devices with edge computing, you won’t want to miss what we have in store. 

Finally, we pivot to look at the commercial side of things. Learn from Dave’s perspective on how to transition from an IT practitioner to roles in product strategy and market messaging, and the evolution of his approach to prioritization. We’ll also discuss how startups can leverage the right team and teammates to make inroads and land big customers, and why IOT has become so mainstreamed into solutions that it's almost invisible. So, buckle up for a thrilling ride through the past, present, and future of Edge and Cloud computing.

Want to scale your edge compute business and learn more? Subscribe here and visit us at https://edgecelsior.com.

Pete Bernard:

When you ask people what Edge Compute is, you get a range of answers Cloud Compute in DevOps, with devices and sensors, the semiconductors outside the data center, including connectivity, AI and a security strategy. It's a stew of technologies that's powering our vehicles, our buildings, our factories and more. It's also filled with fascinating people that are passionate about their tech, their story and their world. I'm your host, Pete Bernard, and the Edge Celsius show makes sense of what Edge Compute is, who's doing it and how it can transform your business and you. So let's get started. But this sounds good, Trust me. I've had some folks with their AirPods on and shouting in their empty office and it doesn't sound good. It's like are you in a men's?

Pete Bernard:

room or something Cool. Well, thanks for joining me. I guess I'll kind of just get started. One of the things I always forget to do is introduce the guests. I usually just start talking and then, like a half hour into it, realize no one's been introduced. Maybe we should start with Dave McCarthy and you sort of give us a little bit of your. Who is Dave McCarthy, for those that don't already know who you are, and kind of your. How did you get from here to there, or from there to here, I guess technically, yeah, I think that's a good place to start.

Dave McCarthy:

So usually I start off by saying that I've been in the IT industry in some form or fashion for about 25 years, but in that time I've worn a bunch of different hats.

Dave McCarthy:

I mean, I started straight out of college just as an IT practitioner.

Dave McCarthy:

I was a system administrator, network engineer, did a lot of racking and stacking of servers, kind of lived that whole life and kind of worked my way up into management of large IT organizations, and I guess that was my first act.

Dave McCarthy:

Somewhere along the way I picked up a desire to not just be somebody that sort of buys and deploys and uses all this technology, but I felt like it was in a good position to help guide product strategy and do market messaging. And so I kind of went into Act 2, which was all around a bunch of different product roles, and there specifically is where I had a lot of focus on embedded software and mobile devices and kind of where the edge computing thing comes into play. And then now my third act I am an industry analyst, so it's kind of an interesting thing in the fact that instead of kind of looking at it from a single company point of view. Now I get this opportunity to see what all the different companies are doing and what some of the problem spaces are for customers that are trying to solve problems, and it all leverages, by the way, all that stuff I did in terms of walking in the shoes and designing products to say like, hey, what advice would you give somebody if they were in those same situations?

Pete Bernard:

Right right.

Dave McCarthy:

So yeah, so right now I'm working as a research vice president at IDC and it's been a great opportunity to both look at kind of infrastructure in general but that whole kind of edge to cloud continuum Right right, yeah, no, that sounds pretty cool.

Pete Bernard:

By the way, I noticed on your LinkedIn, you went to Seton Hall. Are you from New Jersey originally?

Dave McCarthy:

Yeah, I was originally born in New York City, so my family was from the Bronx and then we moved to Westchester County and then eventually New Jersey. So, yeah, I did my undergraduate at Seton Hall. Cool yeah, from.

Pete Bernard:

New Jersey as well, so I have a sensitivity to seeing people's backgrounds that might be from New Jersey, so I saw that. I also noticed you had under your and this is going way back here Seton Hall. You were involved in the pirate radio station.

Dave McCarthy:

Yeah, that was a lot of fun. In fact, I guess if I could have had a second career it might have been in radio or something like that. I mean, I enjoyed, you have a good voice for radio.

Pete Bernard:

I mean, I think I could have a good voice for radio.

Dave McCarthy:

Yeah, I mean I enjoyed it all. I mean, of course, you know I looked at what my peers were doing and they're like, yeah, I can make like $15,000 a year out of college doing radio somewhere in the Midwest. Or you know, there was the technical careers that were kind of calling my name. And so I went that way.

Pete Bernard:

Yeah, I was on the. I went to Boston University and I was a DJ I guess a DJ, whatever they call it on the student radio station there. I had a very late night slot, I don't know. It was like, you know, 10 till two in the morning or something like that. I'm not sure anyone was listening to me actually, but it was fun it was fun to spend some Pink Floyd albums, you know.

Dave McCarthy:

Yeah, that was a unique part of WSOU, and like where it was is because, you know, we had, like our radio tower reached into New York City and all the way down to the Jersey Shore, so we had a pretty good coverage area in terms of, like people that could hear it. So cool, cool, that's good.

Pete Bernard:

Yeah, no. So anyway, moving forward from that, I thought that was interesting. It's always interesting to see sort of like the deep background from folks and sort of you know what their, what their journey has been. We all have journeys and so it's always interesting to see that. So you've been with IDC for a while, so so, yeah, this whole space that you're in I mean, I'm a little biased because I love the whole edge computing space myself but like what's kind of the hot discussions these days around cloud computing and edge computing or just the cloud stuff as well, I notice that you do a lot of overviews and analysis of like AWS and Azure and a bunch of other folks. But is there like a other than, I guess, generative AI maybe?

Pete Bernard:

we talk about AI for five minutes but like what is the? What is the hot? The hot topic these days when you get into these briefings, what is that? What is the hot question?

Dave McCarthy:

Well, you know, so we will. I will respect the let's put the generative AI on the side, but, but to be honest, right, it's everywhere. But I would say, but before that, let's just say you know, six months ago or nine months ago, before that really kind of stole everybody's attention, it's really been around this idea of moving from, like, centralized to decentralized systems. I mean, the concept isn't new, right? And this is kind of where edge computing became kind of interesting. Again. Everybody always talks about the pendulum swing, right. I mean, you can go back in time, right, it was all mainframe, centralized, and then it was all client server, and then web kind of brought it back in again.

Dave McCarthy:

And ultimately, what I'm seeing is people with a desire. I mean, of course they're always trying to solve a business problem and unfortunately, I feel like a lot of times in the industry, when you have a company that's strong in a particular area like cloud, then they're going to tell you that anything and everything that you want to do will work great in the cloud. That's kind of what. Of course they're going to do that because that's what they're there to sell. And it took companies a really it took cloud providers a while I mean some quicker than others to acknowledge that, hey, you know what, not everything runs well in the cloud, and or does it make sense to right?

Dave McCarthy:

So this idea of taking things then back out and putting them in different places, and so I think, from an edge computing perspective, what's been interesting is just the diversity of different kind of deployment types, because we talk about it very like singularly, like hey, there's a cloud and there's an edge, but yeah, but you know, we've learned that. You know, there's there's things within telco, there's things on prem, there's embedded devices, there's a whole spectrum of things, and and ultimately, like people experimenting and trying to decide, like, where to best to put a particular workload, you know, without pigeonholing yourself into thinking it has to be in one place or the other, like that's one thing that I've seen a lot of interest in.

Pete Bernard:

Yeah, I was. I was going to say I think people are realizing that it's sort of like, you know, I think in the UK they say horses for courses but it's sort of like the right solution for the right problem. And you know, one of the things also on the cloud side, as you know, is that, coming from a hyperscaler, probably myself is that cloud is a fairly universal piece of hardware. You know it's kind of fixed hardware, you kind of know what it is. Once you get out into the edge and beyond the data center, the edge becomes very heterogeneous and, like you said, there is no one edge, there's lots of different edge kind of configurations and archetypes and architectures and horsepower and the network connection is going to be plus or minus. So it starts to get a little more tricky to deploy reliable solutions when you have sort of a fixed, maybe a fixed cloud hardware. But the edge hardware can be pretty variable.

Dave McCarthy:

Yeah for sure. Right, I mean that's the. You know, the world would be a really easy place if everything could just be, you know, homogenized into a cloud data center. But you're absolutely right. I mean, and it happens in multiple layers, right? I mean you look at it from a hardware perspective. There's all different hardware platforms out there, especially as you get into these like legacy OT environments, right? So you have hardware standard differences. You've got software standard differences, even things like communication protocols and data formats, like it's. It can be, it can. Everything can feel like a snowflake, and I think that's the major challenge that people are trying to solve today is how do I scale these things when I know I've got kind of on one side all of this heterogeneous kind of environments, but yet I'm trying to apply like a consistent way of managing it, that's. I don't know that anybody has fully solved that problem, but you're starting to see some aspects where that's gotten better.

Pete Bernard:

Yeah, yeah, for sure it's like people are kind of solutionizing a little bit and going up the stack because there's just too many pieces and parts, you know, and too many different ways of putting stuff together to solve problems, that the end of the day, every business has a problem they need to solve. That's why they're doing this stuff and you know, the solution needs to fit inside of sort of an ROI framework and as much as we geek out over the science, experiments and the POCs, at the end of the day the business is like, hey, I have a budget, I have to solve this problem. You know, I don't want to sort of fart around with a bunch of test tubes and chemicals, I just want the, I want the solution. So, yeah, that's one of the, I guess, the challenge and the opportunity. A lot of companies are really going to be making lots of money solving those problems by putting these pieces together.

Pete Bernard:

So and I think these days there's more pieces than ever, like you said, the edge capabilities are really accelerating with new semiconductor platforms and better networks and better workload orchestration and all this other good stuff like people are really investing on improving the kind of capabilities of the edge and so, yeah, you're seeing a lot of companies maybe thinking about how do I run some of it nearby or on-prem, or in the parking lot or you know, or in the hyperscaler, and so I think that's kind of ushering in a whole new way of thinking about not pendulum swinging necessarily but, you know, just kind of moderating and mitigating all edge or all cloud and all that stuff. So it's kind of an interesting time to think about that.

Dave McCarthy:

Yeah, I mean, I think part of it is just because you know we're always chasing the bottleneck, you know, and sometimes the bottleneck is like compute power, sometimes the bottleneck is the network, sometimes the bottleneck is just you know how much space you have to store data in your device and stuff, and so, and that's why I think you know it'll never be anytime somebody feels like they've cracked the code. You know things change, right, the world continues to change, and so I think, from an edge perspective, even more so than in cloud, we need to be kind of looking at today's problem, but then also like what's coming down the road, because you'll need to be able to plan for that and it'll look probably a little bit different a couple years from now than it looks today.

Pete Bernard:

Yeah, no, I like that. There's always going to be a bottleneck somewhere. So, yeah, just as soon as you fix one, there's another one I was going to ask you about. I was reading about CoreWeave the other day. I don't know if you've analyzed them too much, but this idea of a kind of a GPU centric or AI centric cloud. It almost feels like, you know, as silicon has sort of sort of become specialized with CPUs and GPUs and IPUs and NPUs, it's almost like we're starting to see clouds themselves become more specialized for more specialized workloads, which I think is the idea behind CoreWeave. Right, it's really an AI centric cloud powered by Nvidia. You know silicon and do you see that Like? How do you see like the core Weaves of the world relative to the AWS and the Ashes?

Dave McCarthy:

Yeah, so I'm not familiar with them completely, but I am familiar with the idea and there's a few different concepts that are in there, I mean, you know. One is the fact that, you know, Intel for the longest time tried to convince everybody that all you need is a CPU and if you need more capabilities, we'll just pack it into the CPU. Let's get a faster one right.

Dave McCarthy:

Yeah, or you know, and they've done some innovative stuff.

Dave McCarthy:

I mean, don't get me wrong, I'm a fan of Intel yeah, and even they realized right, like they've right, they've branched out, because I think you know kind of you know all these companies that are kind of dominant in their market.

Dave McCarthy:

I mean they have a certain resistance to change because they know what works for them and what's been successful. And so you know, it always takes some other company to kind of come in and sort of disrupt things a little bit, for the bigger and trench ones to say, well, hey, maybe that's a big enough of a trend that I should sort of jump onto it, so so, yeah, so one of this is what I think you touched on, which is, you know, cpus by themselves will not be able to handle all of the sort of complexity of the Applications that people are building today. So you have all of these other sort of co-processor Kind of variants and, of course, the one that's really captured people's mind, or the GPUs for what they're doing with the AI side. But you mentioned that sort of IPU or, like in video, would call it the DPU.

Dave McCarthy:

But, you know again, same idea, right, like you want to offload, like the CPU is good for certain operations, and that's great, but, but ultimately you end up having sort of a collection of Specialized processors, I guess, that are for different things, and so I think that's one thing, which is just this acknowledgement that you know back to the right tool for the job.

Dave McCarthy:

There's going to be different silicon types for different parts of the solution. And then the second part is you know, what we're seeing in cloud more and more is this idea of, like Composability of the infrastructure, or being able to maybe decompose the infrastructure into its elements, such that, you know, you're not just sort of scaling things linearly but you can say, well, you know, I can, you know sort of individually kind of control, the, the components that are in there, and so for something like a, like a GPU or AI centric cloud, it's, it does make sense to me because you're going to find certain workloads that are really kind of tuned for that type of Environment, you know. So I and you see the big cloud providers doing the same thing, right? So there's, you know it's, it's a couple ways to think about it, right. I mean, the, the big clouds are looking at ways of building specialization. But these smaller providers are again kind of like what I was saying earlier. They're sort of creating that disruptive force to make it, make it real.

Pete Bernard:

Yeah, no, I think it's, yeah, I think it's fascinating and we'll see, like just sort of like, as people are decomposing their Workloads to run some more in the cloud, some when the near edge of the network edge, some on-prem, you know, on the far edge, I think we might see people orchestrating workloads to run across different clouds, depending on what the workload is right. So if it's more of a FinOps type process that might run on Azure and then there's a heavy-duty AI that could be running on core Weave and then all that could be sort of orchestrated together across multiple clouds, I don't know, it feels like where that that's sort of like a horizontal, maybe a horizontal Instruction as opposed to an edge to cloud deconstruction of workloads. I'm not sure that makes things more complicated or makes them better, but it's interesting to see. It's interesting to see some innovation there.

Dave McCarthy:

Well, yeah, I mean, and that's you know. I think a lot of what we're seeing today, with everything we just discussed, is happening kind of on that cloud side. So then the question does become so, when you extend that out into the field, how much of it comes with you, how much of it doesn't? You know if you're talking about really small devices, chances are Especially small devices that you know, if you get back into the IOT world, right, like battery-powered devices, I doubt you're gonna see a GPU associated with a battery-powered device, at least not in any significant way in the short term. And and so you know it does Sort of depend back to those like layers of edge. So if you are kind of looking at something like a metro deployment Location, you know, within a city, as opposed to a big hyperscale region, absolutely like that to me fits.

Dave McCarthy:

You know part of the definition of edge and you know we are seeing some interesting configurations in there. But then again, like I said, it has to somewhat scale down to smaller sort of form factors and I think it's still the jury's still out as to how some of these architectures will mature. But the one thing that I'm fairly certain of is that, back to my there isn't just a single edge. I think we'll see more of like what I might refer to as kind of like a multi-tiered kind of scenario where at different points in the, the different points in the architecture, you have like right-sized kind of aggregation, yeah, and eventually it all rolls up to some big cloud thing, because there's gonna be somebody bad in the back office that wants to see yeah, that wants to see the big dashboard, that's sure. Hey, how's everything running?

Pete Bernard:

And you want to store the date over time and all that good stuff. And, yeah, for sure, you know, and look at it. No, I agree with you. I think, though, you know one of the interesting things I don't know if you've followed Getting down to the tiny edge. You know some of the work from the tiny ML foundation and Like all of semiconductor and stuff, so that, like all of his interesting because they're not, there's not a GPU on there, but they do have the arm M55 core on there, so they're doing like this kind of AI model acceleration on like Cortex-M, like MCU based Silicon, so you can do some anomaly detection and other you know image recognition Is the store shelf empty? And stuff like that. So you can, you can run the AI models today on on battery-powered Silicon. It's just a matter of making sure it's the right model with the right use case and as long as the battery lasts a couple of Years could still work, but but I think we'll see AI acceleration across everything from the tiny edge to the you know, cortex-a obviously we have it there already and then to the heavy edge, and you know, yeah, who knows of these kind of you know more vertical clouds Start up.

Pete Bernard:

I was looking at a company called Crucio. I don't know if you've heard of them Friend of mine just pinged me on them last night. They run a cloud service that apparently and I haven't actually dug into this too much so you can put your workloads on their cloud. Their cloud is powered by Excess energy. So you're talking about sort of sustainable cloud operations. So what they claim is that they capture, like the gas flare, these gas flares and the energy from that and they Somehow, you know, capture and transmit that and power these kind of remote data centers off of this excess energy, from an excess energy Grid of some sort. So I thought that was kind of interesting and it feels a little niche. But I mean the whole, the whole topic of sort of, you know, environmental and sustainable Sustainable resources and computing obviously is a huge topic we could talk about. But a crucio, specifically, I was like interesting, like okay, getting the gas flares and turning that into power Electricity is kind of unique. So I don't know, we'll see.

Dave McCarthy:

Yeah, well, I think that's one of the biggest problems that that these companies are going to eventually have to solve. For I mean, they, if you talk to a major cloud provider today and you ask them about like their buildouts into, like other countries and regions, one of the biggest challenges they have is finding enough like renewable energy, yeah, some of these places, yeah, and that that drives some of or inhibits some of their expansion. And, yeah, you know, what's interesting is, you know. You know, when it comes to the edge computing side of things, like you know, I know companies like Schneider Electric is are working on like how to extend some of those benefits to smaller edge footprints.

Dave McCarthy:

But, but the reality is for a lot of companies like you, focus your attention when are the big energy usage and so like, right now, a lot of the attention is on big data centers. It's not so much at what's happening at the edge, but then when you think about how many edge locations and all the, I guess it's a, it's a a brewing problem. Yeah, that I guess, hopefully, like some of the innovation that happens in the big data center, we'll kind of find its way out to the edge, but it doesn't seem to be the. For a lot of companies it's not the biggest problem they're trying to solve. It's like a roadmap issue.

Pete Bernard:

Yeah, and I think there's a lot of. I mean, greenwashing is kind of a strong term, but I think people are trying to be good citizens and make sure they try to be as close to carbon neutral as possible, and so it's easier to do that if you just put all your stuff on a big hyperscaler that has a good carbon neutral program. It's a little harder to do that if you have a lot of heterogeneous pieces and parts that you really aren't even measuring or can measure, like the power consumption on. So you can't really make any claims about carbon neutrality for anything really outside the data center at this point.

Pete Bernard:

But I think it's coming. At some point there'll be some intelligence around power usage and you'll be able to look across the whole taxonomy of your system and kind of understand the maybe the power in and where it's coming from and the power going out. I mean, we're probably pretty far away from that, but ultimately I think people want to be held accountable for the sustainability of their solutions, whether they're completely edge focused or completely cloud focused or whatever. I just think we're probably just not there yet in terms of being able to measure that stuff outside of a data center.

Dave McCarthy:

Now, what I do think is interesting is that there's like, when you get out to the edge and you do have smaller systems, I mean, battery might be part of how their power is supplied, but this is also a place where you see some interesting examples of renewable energy. So I saw, I think, lenovo talking about one of their smaller servers being powered at some sort of remote facility through solar.

Dave McCarthy:

And battery was there as kind of a backup, and they had designed the infrastructure to be efficient enough such that that would work, so they didn't necessarily even need connection to a power grid. Because, again, when you talk about some of these edge locations, there's what I would refer to as kind of IT friendly edge locations, like not everything needs to be ruggedized, and just because it's small and remote doesn't mean it's harsh.

Dave McCarthy:

But then there's all these other examples where it's just like, hey, I'm out in the middle of an oil field somewhere or I'm out in the middle of a forest or I'm doing smart agriculture and there's just there isn't really just power out there or communications in the way that we expect it, and this is where I feel like some of these alternative power sources come into play. I think this is where some of the alternative networking, like private 5G, comes into play, Because these are just kind of like off the beaten path kind of locations right or like LPWA or kind of super power networks.

Pete Bernard:

Also, like you're looking at folks like Qualcomm who talk a lot about going back to generative AI, talk about doing like a lot of AI workloads on the edge using very low power silicon. I mean, they're there, that's their specialty is low power connected silicon where you can run lots of AI vision workloads and things like that with battery powered or solar powered. You know units that are just out there indefinitely and you don't need to. You don't need to crunch a big Azure Stack Edge box or something like that that needs some, needs some DC or something like that. But and then there's also this thing called L-desk. I don't know if you're familiar with that, the long duration energy storage, so that's like the kind of big batteries that can store power for over 10 hours and it's kind of a nascent area. But in a lot of these situations you can take a renewables like a wind or solar and you can feed into an L-desk and then the L-desk even though maybe it's not windy or it's not sunny or whatever, the L-desk will provide sort of consistent power to that remote location over time. So you could have a data center that has an L-desk next to it and has renewables powering it, etc. So, yeah, it's really interesting. I mean, some of this stuff's a little bit kind of years away, but there's a lot of money being put into it and I think, ultimately I think there'll be a lot more awareness and publicity around the kind of sustainability of some of this compute solutions out there.

Pete Bernard:

Because, like I was saying, like when Leonard was saying, you know, don't ask chat GPT what OnePlus One is. I mean, you're burning a whole set of trees just to answer that question. It's not a really good use of resources and I don't think people really realize that. They don't realize that it's not free to do this kind of stuff. It actually costs resources to do it. So, yeah, I think that's an interesting challenge is maybe that's the next bottleneck is the sustainability footprint of your solution, is to make sure that that sort of complies with your local regulations or your corporate governance. Right, that's going to be yet another thing to solve for with these kinds of solutions. So it keeps everyone in business for a long time, I guess.

Dave McCarthy:

Yeah, I completely agree.

Pete Bernard:

Yeah, no, it's fascinating Actually. Well, let me talk about this like give us an idea, like kind of maybe not day in the life, but you know, I've talked to some of these folks that are involved in a lot of the industry analysts stuff and it seems like the amount of sort of engagement and travel is like really tremendous and there's just so much going on. I mean, if you go on a LinkedIn, there's probably like a dozen different things going on a daily basis in 12 different cities around the world. How do you prioritize your time and what you you know? How do you allocate your time and be able to task Swiss without like kind of thrashing?

Dave McCarthy:

Yeah, well, if there was one good thing about the pandemic, it was just the fact that everything was a virtual event.

Dave McCarthy:

So it was possible to be in more than one place at one time.

Dave McCarthy:

If you know, you had multiple screens and those obviously doesn't doesn't exist today, right, and you know, last year I felt like was kind of a transition year, and this year for sure, if I, if I wanted to, I could be on the road pretty much every week, but then I wouldn't get any work done.

Dave McCarthy:

So you know, I think it's for me personally. I mean I get to take a pretty broad view of the market. So when I think of, like, the types of companies that I speak with, I mean sometimes you know it is the cloud providers, and so I try to make sure that I'm at the important cloud provider events or it's with, you know, software companies like VMware and Red Hat, and they have their events, and then there's like the I OEMs, like Dell and HPE and others, and so you know. So for me, you know, one of the things that I always thought was important in my career in general and I think it's incredibly important as an analyst is to try to be as well rounded as you can and you know, get different opinions and perspectives.

Dave McCarthy:

Because, because, ultimately, when you think about these kinds of solutions, usually it's not one company solving the problem, it's usually multiple companies working in coordination to make this happen. And if you can't connect the dots yourself, then you know, then people who are customers of mine won't find my information very useful. So I mean, there's no magic bullet in terms of prioritization. I try to get enough FaceTime with all of the major categories that I think are relevant in this kind of cloud edge infrastructure continuum to make sure that I'm on top of things that are going on. But then I'll also tell you, like, the other thing that I have a sort of a soft spot for are startups, because you know, I think you know, these big companies are, you know they're the big aircraft carriers moving around doing the things and they have huge influence and ability to do it.

Dave McCarthy:

But it's the, you know, it's the startup community that often has, you know, the most interesting innovation and they're the ones you know making the crazy bets and and, to be honest, some of them, some of them, are crazy right, like, I've heard pitches of companies that have a really interesting perspective and I'm like, yeah, that sounds really cool but I can't imagine it will ever get commercialized. But then you know, you find ones that that really have found that white space, that that exists out there, and you know, and those are the ones that you know typically. You know, you can, you can tell. When I talk to companies you're over a year like which ones give you the same pitch Year over year, which means like not much has happened, and then ones what you talked to and you're like holy cow, you accomplished all that in a year. Imagine what you'll do, you know, in another year. And so it's pretty easy to see like which ones are on the, which ones are sort of coasting and which ones are really like gaining some traction, right right.

Pete Bernard:

Yeah, no, totally. I got involved in an organization, a company called conduit venture labs. So for the listeners you can look them up, but they do kind of incubation of a lot of IOT startups and Especially, I think in the IOT in the edge space there's a lot of these. You know they're doing smart ag or they're doing smart power. They're doing something just very unique and and innovative and, like you said, it's it's hard for the larger companies with all the bets and investments to maybe Move that quickly. So yeah, it's great to have a robust kind of startup community there.

Pete Bernard:

And the other thing totally agree with you, it's like I always say the edge is a team sports. When I talk with companies, Thing I ask him is like whose team are you on and which which? Who are the other team teammates that you're taking a dependency on to drive your business? And you know you have to be super clear on on who you can leverage and take dependencies on and vice versa, because if you're on the right team with some great teammates, as you know, good things can happen. So definitely the edge is a place where there's never one company that comes in and solves the problem. It's always a Consortium or collection of folks. So so, yeah, you're right, sort of understanding, like for me it's like not just around the semiconductors or the devices or the networks, it's, it's all the stuff and the software and how it all comes together. That really is the story, and so, yeah, I can imagine I guess I'm making an argument that your job is even more difficult because you have to sort of Understand all these different aspects.

Pete Bernard:

But yeah, I think that's what makes it fascinating and and a great opportunity to for startups to come into the space. If you can pair up with and get on the right team and have the right teammates, even as a startup, you can leverage some pretty experienced channels and technology and Make some big inroads and land some big customers. So it's pretty cool.

Dave McCarthy:

Yeah, and it's a critical decision for a small company because you typically, like every small company, have limited resources. So, while you would love to be able to be everybody's friend, you can either do that you know sort of poorly with lots of people or you could do it really well with one Right, and my advice would typically be, you know, to hit your wagon to somebody that's sort of Understand your value proposition and can help you do that. Because you know, when I think about Some of the even today, like things that I hear about IOT right, like I'm sure we've all seen it somewhere on LinkedIn or somewhere when somebody posts a headline like IOT is dead, or you know IOT was like a big failure or something like that, and I always laugh because I look around and I'm like, everywhere I look there's smart connected devices Generating data that's being used for some purpose, like that's.

Dave McCarthy:

IOT like I think what's happened is IOT has become almost invisible and so much now that it's been like integrated in that people don't think of it separately.

Pete Bernard:

Right. Yeah, it's been mainstreamed as part of a solution, as opposed to a thing in and of itself. Yeah, I totally agree with that.

Dave McCarthy:

Yeah you know, and I think that that's true we're like for a lot of technologies. I mean, I think that's the. You know where we are with some of this AI conversation right now, like we're having to put so much emphasis on it because it's kind of new and different and even though it's existed for a really long time, it hasn't really been applied so ubiquitously all over the place. And and and that's part of the you know the storyline that I'll often tell people when they say well, you know, I did buy into the whole IOT wave and I invested in all this stuff, so I have new data points, and then you know, but they usually have that hangover of but they didn't really get my ROI out of the solution. What happened, you know right, and you know.

Dave McCarthy:

Just to tie this back to edge, I mean, I think the failure that some of those companies had is well, they usually fall into two reasons. One is they may have just not actually had a strong business Problem that they were trying to solve that they could work backwards from, and they just kind of went off and did it because it seemed like the Right thing to do, and that's like a rudderless ship. You know, you'll never. You'll never get to port, you know that way.

Dave McCarthy:

But but if it wasn't for that, even if they knew what they were going to do, you know, it was this issue of like trying to tie together this data in the field with the intelligence in the cloud, and there was just this huge Chasm between those two places that, even if they could make something work in a proof of concept, they could very rarely like scale it to a point that it was useful. And To me, this is where, like the edge computing idea even though, like you know, we put so much effort into saying it's this great new edge thing, but to me it's just, it's an extension of things We've already done. It's like, instead of taking all the data and moving it to where the processing is, if we could take some of that Processing and move it to where the data is, oh, my goodness now.

Dave McCarthy:

It started off as a very disjointed, you know kind of process is like wow, I can do a closed loop system on prem that does everything from collected data to understand the data, to implement action, and then, of course, I'm gonna report back what I did to some centralized source, like it was the missing piece. I think that that Stymied some of those you know early projects and and so if you're embarking on this now and you can take advantage of all these, then I feel like a lot of the inhibitors have been Either minimized or eliminated.

Pete Bernard:

Hmm, yeah, no, I definitely agree with that. I think it's sort of an evolution of the same Concept. But you know, yeah, the tech is better, the networks are better, the chips are better, the software is better, and so, yeah, it's, it's, there's a lot of new ways to sort of solve the problems that people have been trying to solve for a long time and and you're right, gets kind of mainstreamed and it's not as glamorous Because it's just sort of part of the what you do. You know, and I think with AI will see that as well, we're already starting to see that In a number of products, where it's just the way the product works and you know, become sort of an expected outcome, as opposed to some you know, science fiction, cool thing. So, but yeah, no, it's that's the. You know, like you said, there's always another bottleneck and we'll find it, I'm sure. But Let me ask you, dave, if you have any, any closing thoughts or any words of wisdom that you would want to Impart upon us in the edge compute world from your perspective.

Dave McCarthy:

Well, you know we tried to stay away from really spending too much time on the AI thing, but but at IDC, you know we're we're keeping an eye on it and in fact, we're rallying around this term. You know AI, you know everywhere, and so, while it's going to start with big systems in the cloud, it's going to find itself, as we've already seen, you know, out in these edge locations and and really the challenge will just be Back to this centralized versus distributed.

Dave McCarthy:

You need a distributed infrastructure. You also need a distributed data plane and you're going to need these distributed applications. That kind of understand that and so you know. So to me, I'm I'm excited to Look out over the next couple years and see how these things continue to Mature and as all those technologies continue to get better and, and, more importantly, like people understand them better, and they've They've made a few of the the mistakes. Right, touch the hot stove. Okay, I won't do that again. But you know, find something, find something that works and say, hey, let's do more of that. I think we're going to continue to see. You know more and more, you know great examples of it's not just a cool technology thing, but it's actually solving real business problems and creating real business value and ultimately, that's that's what we're here for, right, that's yeah, that's what gets at least, that's what gets me excited. Is that kind of like, hey, how do those, the worlds of business and technology, collide?

Pete Bernard:

Yeah, no, exactly. And then you know, seeing ai as sort of a supercharger or or providing superpowers, you know, to solutions is kind of exciting when we see that, so Cool. Well, dave, thanks again for taking the time. I appreciate it and I'm sure I'll see you around. All right, pete, thanks again, take care Bye. Thanks for joining us today on the Edge Cell Sewer Show. Please subscribe and stay tuned for more and check us out online about how you can scale your Edge compute business.

Exploring Edge Compute and Cloud Computing
Edge and Cloud Computing Challenges
Cloud Infrastructure and Edge Computing
Cloud Providers, Startups, and Collaboration in Edge Computing