AI Everywhere: Powering the New Edge with Cisco - Six Five On The Road

How is AI’s migration to the edge transforming enterprise operations and data management strategies?

‍

From Cisco’s Partner Summit 2025, host Daniel Newman, CEO and Chief Analyst, Futurum, is joined by Cisco's Daniel McGinniss, VP Product Management, Compute, for a conversation on Cisco Unified Edge, exploring how edge AI enables data-center-grade performance anywhere and helps businesses scale secure, manageable AI deployments across enterprise locations.

‍

Key Takeaways Include:

‍

🔹AI’s shift to the edge: Driving forces behind moving compute closer to where data is created, and Cisco Unified Edge’s role in real-world deployments.

‍

🔹Data-center performance at the edge: How Unified Edge brings enterprise-grade reliability and scale to local sites, addressing the challenge of untapped edge data.

‍

🔹Simplifying edge operations: Practical strategies from Cisco for reducing complexity across distributed sites and integrating partner support.

‍

🔹Security embedded from the outset: Approaches for protecting edge AI workloads and minimizing risk without relying on bolt-on solutions.

‍

🔹Accelerating AI production: How Unified Edge helps organizations move beyond pilot projects to achieve “AI everywhere” in practical, scalable ways.

Learn more at Cisco.com.

‍

Watch the full video at sixfivemedia.com, and be sure to subscribe to our YouTube channel, so you never miss an episode.

‍

Or listen to the audio here:

‍

Disclaimer: Six Five On The Road is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

‍

Daniel Newman: Hey everyone. The Six Five is here On The Road at Cisco Partners Summit 2025 in San Diego, California. We've been having a lot of conversations here on the Six Five. It's a big week. Partners, yes, that is what we're all here for. But we know what all the partners and what Cisco is focused on and that of course is AI and we are going to be talking a little bit about that. But we're also talking about some of the announcements the company's made and its compute strategy and what it's doing not only in the data center but at the edge and so much more. Excited to have Danny McGinnis joining me for the first time.

‍

Daniel McGinniss: Thanks for having me. Yeah, I appreciate it.

‍

Daniel Newman: First timer, long time listener though you can, you can say yes even if you don't mean it.

‍

Daniel McGinniss: Yes.

‍

Daniel Newman: All right.

‍

Daniel McGinniss: Long time listener, of course.

‍

Daniel Newman: First of all, you know, it's great to sit down with you as a bit of an infrastructure guy myself, chip guy. I love talking about the build out and what's going on. Cisco's made a lot of investment. It's made a big re-entrance in many ways into compute. Before we dive into what's going on here, I'd love to just kind of get a little bit from you about kind of how that came about, why the company's leaning in there and what's the opportunity you see that's good.

‍

Daniel McGinniss: I mean we've been on this journey in the compute space for a few years now in terms of kind of reinvigorating. So obviously we've had a phenomenal compute portfolio for quite some time. UCS was started 15 years ago, so we've been in the business for a while, and still have a great 50% market share in our blade space. But it wasn't an area that necessarily was growing as quickly just because the industry itself wasn't. This November 2022 ChatGPT thing happens and next thing you know, AI is kind of cool again and infrastructure or AI is cool and infrastructure is really, really cool. And so that spurred what we want to be when we grow up or how we want to navigate this next decade? And I think all the way up through the top we've made some decisions to say we're all in on compute and that's led to us building new form factors, new operating models, new software and really invested heavily in this portfolio and in the space overall in the data center.

‍

Daniel Newman: Yeah. And it's gonna continue to be. I remember there was a time not that long ago, before 2020 when covering chips in the media was considered not cool, like, ah, we don't wanna cover that. Nobody cares about semiconductors. It's like, wow. And then, you know, everything was SaaS, everything was software, everything. And of course they're so symbiotic. All these things tie together so, so importantly. And then of course AI is driving this really big change. It's going to change software, it's changed the data center entirely. It's going to be a great opportunity for growth and it's going to put companies like Cisco in a position where you're going to have to rethink. But of course, with what, 500,000 hands, as they showed on stage here at Partner Summit that are selling Cisco, great opportunity to put the right products into the hands and have Cisco lead this generation too. You know, one of the things with compute though they always talk about is bringing compute closer to where data is. We all know data scale is massive, data centers are huge, energy is a constraint. But we really need more solutions that are saying, hey, let's bring data and infrastructure to the edge and solve problems that way. Yes, Cisco's doing that here. Unified Edge. Talk a little bit about how that's helping make this possible.

‍

Daniel McGinniss: Yeah, so I guess, you know, as we started talking about where do we want to go? Who are we going to be? Obviously we had to build the largest, the most dense AI platforms that you can deliver through Nvidia and others that are generating, you know, offering these semiconductors, these capabilities in the industry. But then it was like, well, wait a second, we're also seeing, if you think about this, where is the actual application going to run when it comes to verticals like retail, financial services, healthcare, they're absolutely reinventing themselves in terms of the capabilities that they need to do closer and closer to the user. And so as we started to say, well, great, we need the large boxes, we need the biggest, we need to make sure we can go into the data centers and suffice all the needs of those customers there. But there's also this opportunity to completely reinvent the way that compute resources are enabled, managed, and rolled out at thousands and thousands of locations. And that's really what Unified Edge was all about.

‍

Daniel Newman: So, you know, companies have so much data, they're trying to decide, right, they're navigating between hybrid, they're navigating between multi, they're navigating between building out their own data centers. Right now we know that cloud scale is incredibly fast. How do you sort of talk to them? About Unified Edge in terms of its ability to give them that performance. And what's the advantages that you're really seeing over saying, hey, just take the, take the workload and do it with a, with a cloud provider.

‍

Daniel McGinniss: Yeah. So I think you have to. It really comes down to two big things. Where the, where the data is actually being generated or housed, and then two, what's the performance characteristics that we're trying to create at the, at the location. So if you think about it. I mean, if you just think about our personal lives, this one is actually one of the things about AI that's closest to home. If you walk into any sort of environment related to healthcare services that you receive, or you walk into a retail location or you go to a fast food restaurant, that experience is changing for you. There's automated tellers, there's new agents that are helping you look at your sunglasses and determine if that's the right sunglasses on your face or how the clothes look on your body. We're seeing surgical robots happening at these medical facilities. And all of that requires that things happen in real time right there. There are cameras watching you and I right now that have all this intelligence that they're collecting about us that they can go make informed decisions from everything from the temperature of this building to is there crowd control, the whole nine yards. But that actually needs to be processed locally in order to do things in real time as well as we don't want to saturate all of the WAN links and the bandwidth that is being. That is going back to some central location. And so it's really changing the needs of an edge location.

‍

Daniel Newman: Yeah. And I think a lot of it comes down to, you know, the definition too. I think there's edges that look more like small data centers.

‍

Daniel McGinniss: Yes.

‍

Daniel Newman: And then there's edges that's more like what you're doing with Unified, which is like a box that's all things inside of a single box that can be put into a, you know, into a retail store's closet that can, you know, be used real time to do augmented experiences for designing a shoe on the fly.

‍

Daniel McGinniss: Absolutely.

‍

Daniel Newman: And of course, then there's more. Hey, we need a small data center that's, you know, in proximity. Like telco edges and stuff.

‍

Daniel McGinniss: That's right.

‍

Daniel Newman: Which are, which are a little bit different, but the setups do get complicated and there is a lot to manage.

‍

Daniel McGinniss: Yes.

‍

Daniel Newman: I mean, how are you guys sort of thinking about that side? Because software matters and that's become a big part of your business.

‍

Daniel McGinniss: Absolutely. So that was really big. As we stepped into this and we thought about what we want to do in this space, there's really, it's interesting. There's a few Personas that are all kind of coming together at one time. You've got your typical infrastructure team that is deploying infrastructure at an edge location or a data center, or thinking about how the cloud resources run. You've got your networking team that's thinking about how the connectivity to and from the edge location looks and how are we going to get packets from our data center back to the edge. And then there's also the application teams that are rethinking what type of workloads they want to run, what type of business outcomes they want to create, what new competitive advantages they can offer to their business. And so we were thinking about this, it's like, all right, well, we had to kind of make sure that we can sort of bring those three different worlds together. And so the platform we put a lot of thought into, yes, we'll rethink the infrastructure itself, but we also have to rethink how this is managed at scale. We've been very, very successful at managing thousands of boxes in a handful of locations, meaning a centralized data center. To do this right, you have to get really, really good at managing a handful of devices in thousands of locations. And so that concept kind of is termed in the industry as fleet management. We brought our same operating model, a tool called Intersight, that we use to manage all of our current platforms, and extended those capabilities to be able to be phenomenal at turning up and managing infrastructure in these thousands of locations. And so it starts from the second the box arrives to the entire lifecycle management of that platform.

‍

Daniel Newman: So you build all this, you bring all this compute to the edge, you have all this data, it creates a lot of surface area, and that surface area creates a lot of risk and security. I know Cisco has made a lot of investments to try to secure the network. Also, you gotta secure all these environments at scale. I mean, talk a little bit about this, because AI, of course, being such a great enabler for productivity and for efficiency and for business growth, it also helps the black hats, it helps those that are trying to cause chaos. How are you thinking about security and how Cisco delivers protection to these customers? A lot of these edges also don't have some of these retail, that their sophistication of it, especially locally, is a lot lower.

‍

Daniel McGinniss: Yeah. So inherently, again, another very, very critical or foundational design concept was you hit on two very important topics One was how do we manage this infrastructure with non skilled tech staff inside your data center? You have some of your most talented teams working there at a retail location. We don't. So we wanted to have it do this in a way that we could drastically reduce truck rolls, highly skilled laborers having to go out to these locations. The general manager at the retail store should be able to service the device without having. And we've successfully done that. The security piece is really important as well. And so we fuse security in at the physical layer to make sure that it's protected so that no one can steal a hard drive or do something that's going to physically damage the device. We've implemented a lot of our best practices in terms of how we provision policies and blueprints in an automated way, but making sure that we're monitoring for things like config drift and having trusted root security and things like that within the box. But then we've also brought in a lot of our additional tools like AI defense, which is really strong at securing the AI models themselves, making sure that we have looking for things like injection prompts and all of the other things that are bad things that are happening within the AI. And then as well as including all of our firewall services as well as our observability services from Splunk and made that all really not just available, but also easy to manage and operate and deploy at scale.

‍

Daniel Newman: Yeah, it's going to be a constant cat and mouse game still. It's just now exponential AI creates so much more opportunity, but it also creates such powerful tools. And it sounds like you guys, I've said for a long time, I think the Splunk acquisition really rounded out the portfolio and of course gave you guys a very compelling data story. You know, because we all think of data sometimes as either, you know, Excel spreadsheets and stuff, or we think about it as unstructured but like log data. I mean there's half the Internet.

‍

Daniel McGinniss: Yes. And different things that's basically machine code that's being put off. Well and then it's about. It's not just so, it's also what you do with that data. I mean how much these tools that you're talking about unlock those capabilities. And so what we try to do here is make sure that we can give you. We don't really want the team. That is, you should be focused on the application that you're trying to run and not trying to think about the complexity of the hardware stack. For a long time we had limits on where we ran workloads based on technology limitations. And so we're almost trying to invert that. Like, you should be able to place the workload where the application wants it to be, not where the technology limits it from being. And I think that's what we've been able to accomplish.

‍

Daniel Newman: Well, when you have a very powerful compute outside of the data center, you can do a lot more.

‍

Daniel McGinniss: You can? Absolutely.

‍

Daniel Newman: It just wasn't a thing. You just couldn't get enough compute horsepower to run these kinds of applications. So we're finally seeing the application demand use cases and the computer compute power at the edge. Finding symmetry.

Daniel McGinniss: Yes. And be able to manage it at scale in a way that you could actually operate in a distributed fashion.

Daniel Newman: Think about what the network can do for that network, enabling that. So look, I think one of the things about AI, maybe where some of the skeptics come in, is just taking pilots to production. You've heard some of those studies out there that say, oh, nobody's getting roi. I've also seen studies that say people are getting a lot of roi. But I mean, to some extent it sounds like what you're trying to do here is very practically take AI and bring it to a whole lot of businesses where AI can bring value. Right. Secure the businesses, optimize the businesses, improve accessibility, improve customer experiences, all these things. Retail is a great way to get customers familiarized. For instance, retail and stores. But manufacturing plants, all these places where an edge makes sense, kind of. What are you seeing there? And how do you put the whole AI everywhere in context?

‍

Daniel McGinniss: Well, I think there's a couple of big elements to it. So there's the traditional large LLMs. Let me build a chatbot to do this inside my company or external. And so that's the large language model side of the house. But then I don't think there's a CEO or a CIO or an AI officer that's working for any of the industries that you just mentioned that isn't thinking about reinventing their edge experience for their customers right now. So you just mentioned it. You've got retail, you've got manufacturing, you've got life sciences and health care. Every single one of those industries is being disrupted by new capabilities. And I think if you aren't able to provide those capabilities to your customer, they most likely have a higher inclination to look at a competitor that, like you said, is designing your shoe. That's a pretty attractive proposition. If I'm walking through a shopping mall and one. One brand has it and one doesn't. And so I think the use cases are quickly being defined. And then there's a team of people that have to think, how do I instantiate that? Or how do I make that actually real in an affordable way? And I think that's one of the things we're really after here.

‍

Daniel Newman: Yeah, it's gonna be exciting to see how this all proliferates. I keep saying we're at the very earliest in terms of enterprise. I mean, we rate so much of what AI is doing in terms of buying these chatbots, these massive LLMs, and how much is being consumed and how much. But the enterprise adoption scale is going to be massive and it's gonna start to take place. Like you said, so many of these production pilots are gonna start to go to scale. And as they go to scale, the amount of compute, the amount of networking, the amount of edge, it's such a great opportunity. It's an exciting time to be helping lead the product strategy.

‍

Daniel McGinniss: Oh, my gosh. I feel so fortunate to be doing the team that we're working with is absolutely phenomenal. There's something new coming at us. I mean, if you're a lifelong learner, there's no better time to be in this. In this space than right now.

‍

Daniel Newman: Every single day I wake up and learn something, Danny.

‍

Daniel McGinniss: It's pretty cool.

‍

Daniel Newman: Well, thank you so much for being part of the Six Five here at Cisco Partner Summit 2025. Let's do it again sometime.

‍

Daniel McGinniss: Absolutely. Thanks for having me.

‍

Daniel Newman: All right, thank you everybody for tuning in to this episode of the Six Five. We are on the road at Cisco Partner Summit 2025 in San Diego. Hit subscribe. Join us for all of our coverage here at the event. And of course, all of the Six Five episodes. But for this one, it's time to say goodbye. I'll see you all later.

‍