Dennis Hoffman, Dell Technologies | MWC25 Barcelona
>> A good afternoon CUBE community and a welcome back to Barcelona, Spain. We're here nearing the close of day three of our four days of live coverage on theCUBE. My name's Savannah Peterson here on the rocket ship with Dave Vellante all week learning the cool stuff. >> Good week? - It's been a great week. We've had so many insightful guests.
We've had a lot of conversations about collaboration and cooperation and I have to remark, there's been a lot of buzz about the telco industry's agility to innovate in this particular technological era, and that's not always what we hear from the tech,- >> Well, we've been talking about this for a few years now and yeah, we haven't had enough Dell this week. >> We absolutely have not had enough Dell. >> Which is a little bit rare for us because we know how much you love Dell. >> I know. I was going to say between the two of us, a lot
of love at theCUBE actually across the entire company. On that note, we're going to welcome our first Dell guest of the week. Dennis, thank you so much for coming to take the time. >> It's a pleasure to be here. Thank you. >> How's the show been going for you so far? >> Great. Awesome. It's a good show. For us, it falls in a great time of our fiscal year.
It's right at the start. It's a way to kick off the year, get organized around opportunities and customers, and it's a highly efficient way to meet the industry. If 13, 30 minute speed dating sessions a day is a great way to do business, but no, it's good. >> So you took over GM telecom, which is of course a new role for you.
>> Yeah. - You were running strategy for a long time. What exactly is your swim lane, if I can ask that? What's your scope? >> It is Dell's success with and approach to telecommunications networks and line of business. I have almost nothing to do with our IT business because that was well covered in doing quite well beforehand. But it is a whole different world when you cross over and get into this stuff, into the network and you need to be able to speak the language, relate to the problems, and so we needed a special vertical markets group to go after it. >> And where do you, is edge part of that scope or where do you draw the line between edge and telco? >> Yeah, great question. Edge, there's many edges.
>> Yeah. - Right. If we say, hey, there's public data centers, there's private company data centers and there's everything else, call that the edge. This business came about because of an observation that telcos own a lot of edge and we weren't going to be successful in edge without being very successful in telecom.
And so I don't have any responsibility for Dell's approach to the enterprise edge. I have a colleague who does that, but telecoms have edges and so it's related in that way. >> So a Lowe's with a small data center, a Lowe's store, that's not your edge. >> Unless they buy it as a service through a Telco.
>> Okay. - Right. So if they decide to get it on- prem from Dell, somebody else, if they decide they're interested in getting that as a service or as on-premise equipment from an operator, then we would be involved. >> But making a windmill, I always use the windmill metaphor.
Intelligent would be your scope, is that right? >> Making a windmill. >> A windmill. I mean it's something at the far edge that is not a sort of traditional IT data center. >> From a go-to-market perspective, no.
Unless the windmill was owned by a Dutch telephone company. >> Okay. Got it. - But from a technology perspective, what technology should we build for edge use cases? There's a great deal of overlap. Products that are not made to go into IT data centers that are professionally cooled and powered, size, operating temperature, very nuanced stuff about the design of hardware and software.
Yeah, I share that with my colleagues that do edge. >> So when, I don't know how it happened, I don't know if it was Michael said, " Dennis, I want you to go do this. " Why you? Not like you're a 30-year telco guy. Maybe that's why.
And how did you learn about this business? >> Well, you had Michael on last year, right? So you'd have to ask Michael why me? I think last guy standing. I have no idea. I do think part of it is in general, there's a belief that part of the job is understanding telecom and the way the industry works. And part of the job is understanding Dell and the way Dell works and what Dell's good at and particularly what Dell is not good at and finding the right fit and the former, understanding telecom, there's a lot of different ways to learn that.
Understanding the nuances of a company and its internal workings, there's no course on that. >> Great point. - So it's finding the right kind of inside baseball versus what is the industry doing as well as not being potentially too caught up in legacy thinking about that industry, and I used to do a lot of this for Joe Tucci at EMC.
>> Right. Right. - And in this industry, I mean, you mentioned legacy thinking. There's a lot of legacy thinking when we think about telco across the board. I can imagine, actually, I'm not going to assume anything.
I'm just going to ask you. You've probably, you've got a lot of customers here on the show floor and a lot of partners I'm sure as well. What is the conversation like as you're formulating and strategizing some of the solutions that you're developing before you're deploying? Is there anxiety around some of these solutions? What are the conversations like? >> Well, there's a lot of exploration. I think there's a maturity curve. There's always a maturity curve, right? So if you say somewhere between we are going cloud- native from Core to Edge to RAN to IT, I don't care. And it is a done deal and we, the architecture people are completely aligned with the people that have to operate it.
They are absolute leaders on the cloud transformation of telecom. It turns out there aren't a great deal of those. All the way to people that are like, look, I just don't think it matters to our business. We're very happy with the way the world works today for us and change is the enemy because most network operators are held to a very high standard in terms of reliability. So on that maturity curve, we try to meet people where they're at and the first piece of that is discovery. Where are you on this journey? I would say there's the middle group.
Let's call these people, the people that are learning and trying to figure this out. That group of people find this to be, this can be anxiety-inducing. >> Yeah. - Particularly as you move from between the domains of architecture and technology, which pretty much everybody understands to the human implications of a change in architecture.
We have very vertical architectures called a radio access network. And by geography we have a single vendor for that radio access network and we rely very heavily on that vendor and we have a RAN engineering team. Okay, now we're going horizontal. I no longer have any of that.
It implies a change in people, a change in process, skill set, and it's the people process part, not the technology part that's really, I think it's the work people have to do. It's the hard part. It's why we choose the word transformation, not because marketing chose it. Trying to apply a disruptive change, not a stroll in the park.
>> So you took that tribal knowledge that you had of Dell and you brought in a lot of expertise. This isn't just Dennis. I mean you've hired well. I've met a number of folks that they've forgotten more than I'll ever know about this industry. >> Yeah, me too. - That's good. You want to hire people that are smarter than you are, right? So when you go back three years, the big conversation was disaggregation of that fossilized stack.
You had some discussion about cloud, you had a lot of discussion about monetizing 5G, which doesn't seem to quite have happened yet. Maybe some private 5G. So when you think about the progression, how do you think about that journey and how long that's going to take? What's your timeline? Hopefully you have some patient capital behind you, so.
>> We have some of the most patient capital. >> Yeah, that's a great call out, Dave. Yeah. >> What are your base assumptions? >> Technology substitution always follows an S-curve. >> Yes. - And when you're on one side of the S-curve,
looking at the mountain, it feels like it'll never happen. And every year that you predict it will, you look silly. And then you hit the knee of the curve and it starts to go and one day you look back on it and it was an overnight success 10 years in the making.
>> Of course, naturally. - So we're in the flat part of that old dive curve right now. >> It feels like we're starting to hit a knee. >> Okay. - Because some of the thought leaders
and opinion leaders in the industry globally and the ones that can really move vendors that have the scale to get people like us and strategic partners of ours like Nokia Ericsson to move, to do something different, to meet their changing needs. They're starting to go and as they go, as the industry has a tier 1, 2, 3 terminology, which of course I wouldn't want if I was not tier one. But many look to those as reference points and as they begin to move, it gives them a sense of confidence that it can be done. It gives them a sense of confidence the vendors have moved and are ready to support the new model.
And so it really feels to me like we're really hitting that knee of the curve. I'm having no conversations with anybody about architecture. I get it. It's disaggregating. I get why it's disaggregating.
I'm excited about it to disaggregate. With the leaders, the conversation, as I said, is now moving into, hey, the people process implications of this are not trivial and can you guys help us with that or who can help us with that? >> Has that changed from three years ago? >> Oh yeah, for sure. - There was a lot of architecture talk three years ago, why, why should we do this? We don't want to move. >> Or we're convinced the architecture's great.
And then if you actually ask, have you spoken to your ops people? No, they'll never do this. Okay. >> Yeah. - So that means for me, I'm never going to sell anything, right? Because they get the veto because they run it. >> Right. - Right. It's like, yeah. And now there's much more conversations with operations folks that are, you can tell that the thinking's been going on.
It's like, hey, how does this job description change? What does the skill set? Are you saying I have to go compete with the hyperscalers for site reliability engineers? That's like a tough pull. Or what does it mean when someone provides all of the network functions, someone different provides the cloud infrastructure when something breaks? I mean I can't tolerate vendor finger pointing. Are you guys going to sit there and go, not my bad man? So I'm down for six hours.
Because if that's the case, we'll never do this. So if you don't start working out collaborative support agreements in the ecosystem, mature the software, hardware interface and importantly behave as one company, as an ecosystem when I call, it won't happen. >> I want to ask you about cloud and draw on your experience in IT. We saw with the cloud players in the last decade, the traditional companies like Dell and others couldn't move fast enough to catch up to this big sucking sound that was cloud.
We saw that. It's just like, okay, we'll get there eventually. And finally you got the operating model down. It's interesting. With AI,
I think it's much different, but,- >> Agree. >> In telco, the cloud players are clearly a threat to a lot of the telco folks. My premise is they have to move fast or else, they got to figure out this people and process or else the cloud guys could repeat. Is that a viable premise or are the dynamics different in telco? >> I think in part it's viable, but I think the part that differs is the collective investment of the world's telecommunications companies in mobile infrastructure and in-ground infrastructure.
Even the hyperscalers, if you took their complete cash balances and ability to take on debt, that's a big hurdle to kind of replicate what is today the edge. Michael made this point, I don't know if you talked about it with them last year, but in his keynote last year, he said, "The edge is where humans meet data. It's like everybody's backyard." >> It is. It is. >> And you're in everybody's backyard right now and every small business that's passed by cable, head ends and fiber.
And so I think while there are many edges, the ultimate edge is going to be extremely hard for anybody to replicate from a physical infrastructure perspective. Where the premise is dead on is the operating model. I mean, clouds not a place. It's a way, it's way a of doing stuff. There's hybrid clouds, and multi-clouds, and private clouds, and public clouds, and edge clouds.
>> I mean, it's a place, it's just a bunch of servers in various parts of the country. >> Yes. And at various stages of a network. Could be centralized in a big professional data center, could be at a base of a cell phone tower in Montana. Right. >> Right. - And that's where I think it all's going
to depend on where does the action happen? and I think the same is true of AI, right, because basically AI follows the data. >> Data gravity. - Yeah. >> But so the,- - Well, it's got to meet it where it is. >> And the infrastructure has gravity.
For the telcos you're saying that buys them time to figure out the people and process piece. >> It buys them an advantage starting point from a real estate perspective, particularly depending on the security and latency needs. And ultimately the data gravity is almost everything. The data that feeds the AI, where does it live or where will it be created because it's always cheaper to move compute once to data than data all the time to compute. And we're at the dawn of that. We have no idea. >> Very early 2% of PCs, AI PCs.
I know you're talking about AI devices in our hands and our workplaces and everything else. Couldn't agree with you more. So excited to see what happens as we really ramp up here.
You mentioned that you feel like we are at that meme moment. We are about to start to see this. You're not having the same conversations you were probably getting a little fatigued of about architecture.
What's the conversation you think really needs to move forward to see greater adoption of some of your solutions? >> Oh. I mean, if you'd wave a magic wand. >> Yes. That is exactly what I'm asking. >> A mentor of mine years ago told me, " Businesses make decisions for one of three reasons.
Make money, save money and stay out of jail. " Revenue, cost, and risk. >> Facts. - The issue with network disaggregation
other than it's a cool sounding phrase that really is helpful at cocktail parties, is the fact that we've been unable to tell a revenue growth story. We have had a complex cost saving story and there's no way to make this less risk. It represents architectural change, it is risk. So if you can't,- >> Oh, that's good point. >> I mean, if we wave a magic wand, one make money story, one reason that here's an application you can run in the network to grow your revenue as a telco that requires you to have a modern network. And I think AI is tickling that at the moment, but it's not entirely, as Uber was to 4G LTE services and acted as an analogy for how to disrupt an existing industry with a new technology.
We haven't found that for 5G or 6G enterprise with network modernization. >> That so called killer app. I mean, applying AI to the network might save them some money.
Maybe it squeezes a little bit but. >> Yeah. It's kind of that slam dunk. But do you have any predictions for where we might find that? >> Oh, that's not smart. Oh, where we might find it. Yeah. The proving ground most people have been looking at
is distributed data intensity, like video processing, things, right, that clearly are like, that's just too big and too expensive for me to keep traversing networks with. The needle in the haystack has been, if I have on-premises, let's say a retailer for video enabled fraud prevention. The retailer's big enough to stick a box in a closet and run an app because if you detect fraud, you want real time somebody to intercept the human before they leave the building. >> Yep. - Right. Versus run it back even to the nearest cell tower. If your EV sees a shadow go across the road and must quickly determine whether it's the shadow of a tree or a deer.
>> Yeah, or a human. - It could do that kind of in the car. I'd like it present. So we found we know a lot of stuff that belongs in the network Core. It's obvious. A lot of stuff that belongs on the actual enterprise edge interface. Finding that one notch back has been hard.
But some of these ideas around video and other types of highly intense media are the ones that really seem to be most promising. AI's the wild card. Right. What retrieval, augmented generative AI application might one day exist to take a spinning feed from right here in the cell tower where nobody's like, well, the economics of putting it on- premises are superior or the data is so vital to me, I really don't want it to be in the network. >> Right. - And believe me, there's a lot of us suppliers and operators that are pretty actively scouring for that.
>> Yeah. And who knows what agents bring on top of that. >> Did you just invoke agentic AI? >> I did. I did. - Yeah. That's good. >> But I'm wondering as the leading question, what this world thinks about that. I mean, the enterprise is confused. Is it even on the lips of telcos? >> Oh, for sure. I mean, again, I'm not certain any of us truly understand it's implication, how quick moving it is, but absolutely because at the end of the day, there is a certain computing intensity to it.
There might be a data storage intensity to it, and that means there might be some economies of scale to the infrastructure. And if 15 dry cleaners may say that's way better if I get this as a service from Verizon than if I try to do this myself because it's not as simple as an AI PC. There's something that's got a little data center intensity to it, but it's the kind of app I can't run all the way back to the network Core and run out of AWS. >> Yeah, and its risk associated with it. >> Always. - Right. >> Oh, yeah. - I mean, a security standpoint.
>> Yeah. Well, yeah, and you got a lot of data in that scenario. There's a lot of X factors that can go a little haywire. >> It's early, super early.
Although this one is kind of going at a half-life of its quick. >> Yeah. - Well, your point about the S-curve, it's definitely feels different this year than it did certainly,- >> Very much so. >> Two years ago. - I think more so than last year. >> More so than last year. - I think people were hopeful
last year or at least talking about hypotheticals. I think we are starting to see some action here. I think we're starting to see that the beginning of that avalanche, since you're both snow fans, that's about to happen when we have adoption, but it's also about all the players coming together. How powerful are some of the partnerships that you have in driving this forward? I mean, in telco it's one of the most important things. >> It is. I mean, this is an ecosystem problem, but the world is used to buying from one person, so the partnership's everything.
If we as an ecosystem cannot convince the operators that we can produce something that's cost competitive, equally or better performant, and importantly with every bit that support infrastructure in place, again, it's going to be hard to move. I definitely try to think of and make sure our team thinks of we are in the middle of a two-sided market. There's a group of technology vendors over here. We are a relatively low-in-the- stack horizontal platform component, and our role as a potential ecosystem enabler or platform starter facing the partner community is probably every bit as important as our engagement with the customer. So we apply almost as much resource to strategic alliances and partnerships as we do to customers.
And we have a ruthless definition of strategic partnership. I understand your strategy, you understand my strategy. We see the overlap and we apply a resource against that.
Otherwise, there Barney deals and strategic partnership is like nothing. >> And you're the throat to choke. >> In some cases, we are not. We're like an advisor. We've seen the movie before, to both parties. I mean, a lot of the people that are historic telecommunications vendors don't actually have a great deal of experience with cloud transformation either. >> Absolutely. I was thinking that.
Yeah, that's where the breadth of what Dell does across the board really plays in here. I'm curious, given your unique position to see a lot that's happening in this space at a variety of different scales, what excites you most, well, perhaps as a business, but also I just want to ask you personally about the telco industry right now? >> It's like electricity, right? We take it for granted. It's everywhere. I mean, I think what excites me about it is its power to help to change kind of the way we live.
I mean, most of us in this industry are kind of technology geeks and really new stuff. >> Yeah. - And this is an underpinning technology for stuff that was brandy new not long ago. That is vital and we take for granted today and the pace of innovation is quickening. So we've talked about the various eye candy, AR, VR, drones and all this kind of stuff for a while now, but there's just little examples here and there, right, Meta Ray-Ban Wayfarers. >> Yep. That's come up a couple times on the show this week.
>> Yes, it sure has. - I mean, it's like,- >> Yeah. I know, it's an interesting,- >> Got to get me some. >> I know. No, we need CUBE Meta Ray-Ban glasses. >> You should. - Might have to do it.
>> They're really cool. Just, it's not like you're wearing a first gen Oculus headset. You're wearing,- >> Or Google Glass for that matter. >> Or Google Glass.
A pair of sunglasses or transitions lenses, but you're wearing them and you're walking around and you're like, what's that? Today, it's telling me that through my phone. >> Yeah. - At one point the Apple Watch needed to have the phone. Now it doesn't. >> Right. - The pace of that stuff. Every day helping people connect, education information, that's exciting.
And we take totally for granted that this industry underpins all of it because of how they got overtopped historically. And I think that little nugget of, which we've talked around in different ways, how not to get overtopped, how to find the killer app, how to play a role, charge a toll for the edge real estate is both as a business and as somebody who's trying to figure this problem out, that's exciting. >> And even if they do get overtopped, it's still a decent business.
>> It is. Right. Look, it just has implications for, are you Walmart or Tiffany's, right? >> Yeah. Right. - Yeah. >> There's always the, I don't want to be known as dumb pipe. Well being dumb pipe with the right cost structure could be called money printing.
>> Yeah. - Yes. >> Yeah, exactly. Some of those businesses are the strongest and,- >> Yeah. Or saying, I really wish to be a programmable network with APIs and compete head on with some of the hyperscalers in my sector. Also, cool if you pull it off, but don't get stuck in the middle.
Don't have the cost structure of one. >> Yeah. - And the revenue stream of the other. And I think we're seeing people really kind of rationalizing that, coming to grips with that.
>> I think,- - What are we really good at? Let's go. >> Yeah. - Yeah. I think you're absolutely right, and I appreciate that you brought up the kind of under hyped nature or that taken for granted aspect because it's mission-critical, especially as our workloads only accelerate with AI and with everything else that's going on. This has been awesome. Dennis, I have one final question for you.
>> Okay. - What do you hope to be able to say at next MWC that you can't yet say today? >> That the number of people who are leading on that maturity curve have grown at a multiple, to basically be able to put data behind we hit the knee, right? That it's not Verizon, AT&T doing a deal, their deal with Ericsson and Dell. Vodafone kind of starting to roll out and the upstarts of Rakuten and Boost Mobile or what we used to call Dish. >> Dish. Yeah. - That it's now, oh, there's actually actually 15. There were five last year. >> Yeah. And they're disrupting.
>> Right. - And yeah. >> It's a 3X multiple. The snowball has started avalanche Ski person. >> Yes. - There we go. >> Awesome. Dennis, thank you so much
for taking the time out of your very busy schedule this week. >> Thank you for the opportunity. >> Really appreciate it. - Yeah. >> A joy for us. And thank you Dave. That was a fun one. And thank all of you for tuning in to our fantastic power packed four days of live coverage here at Mobile World Congress in Barcelona.
My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.
2025-03-07 08:22
