52. Servers Are Still Relevant in the Age of Cloud with HPE – Tech Field Day Spotlight Series

Although we live in a world of software, server hardware still matters. From the data center to the cloud to the edge, this episode of the Tech Field Day podcast features Scott Schafer, a VP and Chief technologist at HPE, discussing the evolution of the server with Jack Poller Von F and myself, Steven fst servers really are still relevant and maybe more relevant than ever. Welcome to the Tech Field, a podcast where we bring together a group of it technical experts to discuss a single idea about key concepts in the industry.

This podcast features a variety of perspectives from members of the Tech Field Day delegate community, and is often recorded in association with one of our events. Tech Field Day is part of the Futurum Group, and this podcast is also published on our sister company Site Techstrong tv. On this episode, presented by HPE, we're discussing server hardware.

Is it still relevant in this age of cloud and platforms and software eating the world? Before that discussion though, let's meet who's on the panel today. Hi, I'm Jack Poller.

I am founder and principal analyst with Paradigm Technica. We are an industry analyst and research firm, uh, focusing on AI and cybersecurity. Hi, I am Vung Fa, I'm within generation.

I'm a senior solutions architect. Uh, we are, you know, solutions provider. We provide hardware, software solutions for maintenance customers, public and private.

And Howie. I'm, uh, Scott Schafer. I'm the vice president of our, uh, advanced development group here and the chief technologist for our compute organization.

That's the group in, uh, Hewlett Packard Enterprise that develops our ProLiant and synergy and Superdome class servers. It's great to be here. And I'm Steven FoST, organizer of the Tech Field Day event series, including, uh, the special, uh, event that we're holding with h HP's ProLiant team today as you are listening to this podcast.

So let's start off, uh, by talking to, uh, the team here. You know, we live in the world of cloud. We live in the world of everything as a service.

Why does somebody really care about server hardware anymore, Scott? Well, I mean, uh, obviously the software's gotta run somewhere, and so we care. Um, but I say that somewhere part because I think that matters, the pro the location actually really matters.

Um, we speak to, or I end up speaking to a lot of customers where that location is critical, and especially being really close to the data, uh, being able to, uh, access whatever you need whenever you need it to either run your AI or to build your services or to build whatever, uh, business solutions you're trying to, uh, to build there. Uh, being able to clo being close to the data is, is very critical. And, uh, you know, I think we know that, uh, it's not, uh, cheap, uh, or, uh, effective to backhaul all that to cloud all the time.

Uh, nor do many customers kind of want to have their proprietary data up in, uh, someone else's service. And so, uh, yeah, we, we definitely see a huge and continued interest in growing interest in, uh, making sure that, uh, the customer has on-prem server hardware for whatever they need. Yeah, so it, it's, in part, it's a manifest destiny.

You have control of your environment fully. You have control of the, how long you use it, why you use it for everything. And it's, it's at your control and, and your disposal.

You don't have to be at anyone's whim. You have better economics, you have better control. So why not, you know, you have to invest in yourself and not be at someone's whim.

So that's, that's important in my opinion. And, and some topologies don't allow a public cloud to be honest, or any other hyperscaler. Well, you know, you talk about location and you talk about sort of, you know, control and, you know, long, you said, why not?

And I used to look at the world and say, um, why invest in the hardware? Right? There was, it's, uh, a huge capital expense, and a lot of what companies were doing was, um, migrating their capital expense to opex, uh, for and moving to the cloud.

And that changed the budgeting equation. And then we sort of sort of divided the world into, um, companies that were born in the cloud that started out with the cloud footprint and would never have anything on-premises. And then you had the older companies that had, uh, an on-premises footprint that they would always keep running because the cost of sort of migrating the entire environment to the cloud didn't make sense.

But I think the age of AI is changing that. And you know, Scott, you mentioned that need to, um, have your compute close to your data rather than VOing the, the data transport cost, which is both a time and monetary cost. And so I think as more and more companies start doing and figuring out what they can do with AI beyond just simple chat bots, um, there's gonna be a lot more edge computing and will need for not, you know, for high density, high power compute in these weird remote locations.

So I think in that case, you know, sort of servers really do matter now, and it's no longer sort of a why do we care? It's of course we care. Well, yeah, I agree with you.

Workflow is important. Understanding our data, understanding how we consume our data, what, what our con competitors do with their data. You may be just a service provider, and that's important for its own mission, but when it comes down to it, even your service provider services has competition.

And if other people, you know, privacy issues, there's all those reasons are reasons to keep, uh, hardware relevant in your space, whether it be your own data center, control that at a co-location, but it's still yours. And you have, again, I'll re I'll re echo the fact that it, it's, you have control and then, and that's, it may sound crazy, but it's not. It's just something you have to be cognizant of it and control and cost control and access, et cetera.

Well, then we see, uh, plenty of, uh, systems. Now, it's funny, uh, you mentioned the edge, right? Because I mean, I'm, I I say funny because everybody defines edge slightly differently.

Um, you know, I talk to some cloud providers and the edge is a colo in your city. Uh, okay, that's not what I mean. What I mean at the edge, and by the way, but I don't mean all the way out to your house.

That's a little too far, uh, for us. Um, when we talk about edge from a compute perspective, I mean, we're seeing, um, servers go in at retail locations, multiple servers now at retail locations. We see servers going in at distribution centers, distribution warehouses.

We see, um, uh, servers going in at, at manufacturing, uh, locations. We see them at, um, at sites like, uh, out at oil rigs, we see them out at, uh, uh, you know, really just, uh, base of cell towers. Obviously.

It's like you, you see them in all kinds of, uh, locations, uh, again, almost exclusively because you need the proximity, uh, to the data, right? Again, just backhauling all of it is, is too much. Um, but yet you absolutely want to process it.

We're seeing lots and lots of different kinds of sensors. I'm just gonna use the word sensor. You can insert whatever your favorite kind is there, uh, from cameras to voice to temperature, there's all kinds of stuff.

But we're seeing all kinds of sensors at these locations that our customers really have to be able to, uh, uh, collect all the data, collect all that telemetry, analyze it, and then act on it locally. And that can only be done, you know, with systems right there. Well, you hit anything in on the nose, Scott, telemetry data.

Those are, those are pervasive in our society demands and our consumption of information. When you go somewhere, you ask a question at a point of sale, you want to know a lot of information. Where does the information live?

If it's too far away, it's in or in somebody's foul box or in a, in a, you know, granite Mountain somewhere, it's gonna be something that you, you want access to, right? And, and people don't have the patience. They want the information yesterday.

They want it immediately at their fingertips disposal. So AI is empowering that, uh, 'cause that does the analytics. We have smarter systems at the edge, which actually enable customer, uh, customer experience, cust, you know, endpoint user experience so they can get what they want to do and, and make a better decision.

'cause we're all about making better decisions and we, how can we do it? We don't have access to the data to make information happen. So yeah, I echo that.

I think, I think part of that is because we're looking at different use cases with different desires of business outcomes, right? And if you, if you sort of roll back, I don't know, five, 10 years, the business outcomes and the use cases were very, very simple. It was virtualization and it was just running standard compute.

And at that point, you know, virtualization essentially said that your hardware platform was generic, right? It was, it was a commodity piece. And ultimately it really didn't matter what you were running on.

Now we've gone beyond sort of virtualization and to another use case where what you're running on does matter because of all these other things we're talking about, the ability to have different sensors that might not be able to do it, the locality of data, the latency, the cost, the speed, the environmental conditions of the edge, you know, this generic sort of place that's not really generic at all. It's very unique for every very each customer and different business outcomes, different workflows. So in this case, you know, as you start moving beyond sort of that, uh, generic use case of running virtualization, then the hardware really does start to matter again.

Yeah, I think that's the, the key point here is that when people think, oh, uh, hardware doesn't matter, what they're thinking of is this sort of, um, uh, I don't know, commodity space where you've got, uh, you know, generic hardware, whether it's virtualization or, you know, uh, containerized, you know, containerization, uh, and they're imagining this world where you just stack up, I dunno, one U blade servers as, as deep as possible. And, and they are generic. And, and honestly, those are generic to some extent if, if it's used for that use case.

But what we look at, when we look at the server line of any large provider, especially the one we're talking to here, HPE, um, it's not generic servers, is it? I mean, these things are purpose built. You have different servers for different purposes.

And yes, one of the purposes is stack 'em tall and have 'em be identical. But many of the other purposes are, you know, memory intensive, io intensive, uh, accelerator, intensive edge servers, et cetera. These servers are very different from each other.

And yet, of course, they're all built with, um, many of the same components, but put together in meaningfully different ways. And, and Scott, that to me is what makes this world of servers still so interesting in that you look at them and you're like, okay, so they have, they, they look similar, you know, from a block diagram perspective, but look what they did there. They made this one have, you know, all sorts of memory, or this one can handle all sorts of storage.

And those two servers really are different, aren't they? Well, you hit a, you hit a big paradigm, but catch 22 of software driving hardware, hardware driving software, and use cases on top of that. And when you get with a manufacturer that is understanding that balance, that equation, and making accelerations on both sides possible, that's amazing.

And, and, and that's, that's, and that's the, the value that the, that, you know, hill Packard does. They'll, they'll listen to the customer, they'll see what the use case is. CXL is not dead.

CXL is is, it's still there and it's gonna be more and more important. So it's, it's, it's in, it's, you'll see some new development with that shortly. And, uh, I'll leave it at that.

Scott, you're gonna say something? I'm sorry, go ahead. Yeah, I was gonna say, you know, we, we absolutely are building, you know, some servers optimized for purpose.

You know, we, we recently introduced one that's designed for kinda these edge locations. And you say, what, what makes a server unique for an edge location? Why, you know, wasn't it just a regular thing you just, you know, put it on?

It's like, no, we built one that, uh, that can be, uh, mounted to a wall. It can mount it under a desk. And yes, you can absolutely put it in, in a rack kind of form factor if that's what you want.

But having different mounting options, as simple as, as simple as that is, is something that we definitely heard from customers and said, Hey, I've got all kinds of different locations. This has to go, uh, uh, enable me to do that. Um, we have the ability to, uh, to have a filter physically on, on the front.

Um, I was, uh, I was working with one customer and, uh, they sent me a picture of where the server was, the server's in the break room for the employees at the, at the retail location. And so I was like, huh, what's in the atmosphere there? You know, I don't, I'm not sure.

Uh, the, and so you don't want that kind of stuff, uh, in your server. So we have, you know, filters that go on the front and, and make sure that they're, uh, protected, which is not something you worry about in the core data center, right? A data center, you don't care, you know, it's all filtered specific care, but, you know, on the edge, it's, it's not, right.

So you Think you don't manage that. But let me tell you, I cleaned the filter on my ProLiant, uh, ML one 10 the other day. It's been in a nice clean environment.

That filter was filthy. Yeah, you got a lot of dust in the air usually, right? But then, uh, uh, but then things like physical security, right?

I mean, we wanna make sure that somebody can't physically tamper with or, or take items out of it. I mean, you know, again, this is one of those things I think a lot of people don't think of. 'cause they're like, I physically control my data center.

Uh, the racks are closed, that's great. But when you're out at the edge, right? Think about it, you got that hourly wage employee going along there, you just do, you need to actually secure these things physically, um, in a way that, uh, perhaps we haven't.

So that, that's like purpose built. Um, but outside of even, I mean, I love talking about edge, we talk about it all day. But, um, outside of that, even in the core data center, and we're doing things to optimize, especially around efficiency.

If I, if I had a, uh, of the buzzword for what I think we're doing now and in the future, it's gonna be about driving efficiency. In the past, we've had other kind of things that we optimize for, like, uh, density. Um, but I think we're, that's, that's not what we're hearing, right?

We're not hearing from most customers. They need that anymore. What they want is much more efficiency from, especially from an energy usage perspective.

And so we've been doing a lot to ensure that, um, you can pick good efficient options. If you need the absolute highest performance possible, I can give you that. Um, you know, and you don't care how much power it uses, that's great.

But if you do care and you need more, uh, optimized for energy efficiency, power efficiency, you know, we, we'd give you that too as well. Well, you, you talk about edge servers, Scott, you know, I've been in those single closet areas as they, they call a data center. And the airflow is, uh, is, is not what you expect.

You might not be able to breathe to be, uh, frank. And yes, double is a HEPA filter. So you, you, you definitely have some environmental challenges, not just, not just the compute challenges, not just the use case challenges for driving the software.

Telcos have that Because we, we, the server, you know, was the DL 1 45, gen 11, uh, can operate up up to 55 c. That's, that's a hundred. And what is that?

130 Fahrenheit. So, because as you said, yeah, you go in these environments, they don't have server, uh, sorry, data center class air handling, right? They got whatever air conditioning's in there, and sometimes the back isn't air conditioned at all.

Your ProLiant doubles as a HEPA filter, like I said, right, Steven? Exactly. But you, you, you know, you talked about efficiency though, and it's, there's a number of considerations for efficiency.

You talked about power efficiency, and I think we've been very focused about that recently, um, with, you know, uh, the amount of heat and power these things consume and generate. Um, I think, you know, the other part of the efficiency is things like how do you manage your fleet of servers, right? And I think, um, it's management and it's how do you secure it?

And physical security is one. And, and the securing the firmware and guaranteeing the, the, the trust factor of your data center. Um, you know, you mentioned, you know, physical security and, you know, a minimum wage employee.

When you look at these servers that are in, uh, retail location, one of my favorite retail locations is at grocery stores, supermarkets, right? Your server ends up in, there's no server closet, there's no data center. It's the only office in the building is the manager's office.

And it sits there with their paperwork with everybody going in and out all the time. And no cooling, no heating, uh, coffee spilled on it, all of those types things. So anybody can come in and access it, can put a USB drive in it if it accepts USB drives, right?

Can plug in a keyboard and mouse and screw around with it, pull disk drives out, all of those types of things. But if that's the environment it lives in, how do you ensure the security of the boot firmware or the operating system, all of those things. So I think that's another consideration where, uh, a lot of work you're doing, right.

I think you're, I think you're redefining DLP there, Jack. It's the world we live in today. It's a completely different world than a, a machine stuck in the data center.

A hundred percent. Yeah. I mean, we spent a lot of energy with these, uh, proline gen twelves where, uh, we've introduced a new ilo, new IO seven.

It's an asic uh, design that we do ourselves. I think a lot of people don't know that we, we develop our own ASIC care. Um, we don't buy it from the market e everyone else does that.

I know of every other vendor buys theirs, which is fine. Um, but we develop our own because we wanna make sure that it has, I always say everything I need and nothing I don't, right? I, I don't wanna pay for anything I'm not using.

So it's got everything that we need in there, including our new, uh, secure enclave, which is a new secure vault built into the ASIC itself. It's isolated from the rest of the asic from a, uh, from an access perspective. And it has its own memory and its own storage, completely separate, but it's inside the package.

So nobody can come along with a chip clip and s and clip on and, and extract any information. We use it to store all of the passwords, the keys, the configuration parameters. So none of that is exposed.

And then we use that as well to allow us to make sure that all of the firmware that on that, that is on that platform, first of all, it's encrypted, it's managed by us, and it cannot be tampered with or modified in any way. And if it is, we instantly detect it. And depending upon what it is, we either shut the system down and alert you, all those kinds of good things to let you know something's not going right.

Uh, but that capability is rooted all the way down into the ASIC itself. We call it silicon root of trust, because the, it's built into the ASIC hardware, so there can be no tampering with the supply chain. Nobody messing around with the server in transit and putting a different firm on.

You can't do any of that because the, our, uh, HP's, uh, protection is built right into the actual ASIC itself. The very first bit powers on. And the very first thing it does is load a little thing, you know, we call the boot loader.

That boot loader is protected by, uh, uh, a digital signature in the asic. And so if, if it doesn't match, it doesn't load, just flat out doesn't load. Uh, and so then we chain the tru up from there, right?

The bootloader then checks the next piece and ensures it's signed with HP's keys, and off we go, uh, and build up that whole kind of, that whole root of trust that it's rooted all the way down into the silicon. Um, certainly something that's unique, uh, to HPE. Yeah, I think a lot of, a lot of people think of ilo, um, and, and other, uh, management capabilities of servers as being sort of, um, KVM, right?

I want be able to reboot the thing. I wanna be able to see what's on the screen, et cetera. Um, but IO ILOs a lot more than that because you can also do things like, you can check firmware revisions, you can update, um, the firmware, you can update the software.

You can actually, uh, reconfigure the whole server from there. It's really, uh, you know, hands-off management and, you know, again, I I, I have a ProLiant. Um, I'm still only at ILO five, but, um, but even with, even with that old thing, right?

I can do a lot, um, without having to have hands-on, uh, on that server. In fact, I don't, I don't see a need to have a monitor on the thing, because you can do so much right there. And I will.

But also to, to that point, adding to that conversation, the, the calm management that HP provides is something that's, no one really talks about a lot in the field. It's new, relatively new, but allows you to do upgrade lifecycle management for non-data center for, for a disorganized traditionally, uh, topology. And you have the ability for air gap.

You have the ability for remote sites all from one central centralized location. And that gives you the ability to be agile and flexible to the changing business LA landscape and how people can consume compute. Like you mentioned Jack, uh, you know, you need power at the edge.

You know, Scott mentioned that as well. And then there's AI use use case POS systems. And I think that shows the evolution of computing that, hello, this is gen 12.

You know, they've been at it for a while, right? So hopefully some lessons have been learned and, uh, and applied and there's a good strategic direction. And I see some of that, not all of it, but it's getting there, right?

Yeah, I think, I think it's getting there. And I think what we're seeing now is, um, a change, again, as I was talking about earlier, a change in the use case from I need generic compute, and I just need, uh, you know, my requirements are processor and memory and some storage to, uh, a much more comprehensive list of, uh, capabilities that I'm going to use, which is, you know, security, remote management, um, power efficiency, space efficiency, um, environmental conditions that are different, locations that are different hands-on, hands-off management. That's different than it used to be.

And the, and I think we're also gonna see now, um, capabilities that will help us do, uh, a lot more automated management as we get bringing AI into, you know, our automation capabilities to reconfigure things on the fly depending on use cases and time of day and needs, et cetera. Um, so it's gonna be pretty interesting what's moving in the future for servers in general. And, you know, for me, Steven, this sort of goes back to answer your question of does do servers matter anymore?

And I think I would've said no, uh, a year ago. And today, I say absolutely it does. Well, and then of course, as we've kind of alluded to here, you know, in this world of AI servers, um, there's an entire universe of, um, accelerators and expansion cards.

There's disaggregation with CXL. There's incredible capabilities in terms of memory and storage expansion. And all of this is leading us to a whole new world where not only do do servers and server technologies matter, but server technologies are almost the defining thing that allows us to move into the future with some of these AI servers.

You know, you look at what's being delivered now, um, to hyperscalers and to the AI factories as well as to end users who are trying to do inferencing so and so on. Um, we've got an unprecedented level of power, uh, requirements and power. Of course, every kid, every watt you put into that server comes out as heat.

So you've gotta deal with, uh, with heat, we're looking at incredible liquid cooling technologies. We're looking at, um, you know, just absolutely, um, amazing demands being put on the hardware. And Scott, I know, you know, the teams that you have at HPE are dealing with this almost existential question of how can I possibly build a server that contains this fireball of a processor, uh, uh, a stack?

How do I build a server that, that can support that? Yeah, absolutely. You know, when we talk, most customers that we talk to, they, they really, um, they rely on air cooling.

I mean, I, I love the liquid cooling stuff. It's really fun. Um, and our HPC customers use it all the time, which is great, but most of the enterprise and, and certainly the, uh, uh, medium sized businesses, lot of colos just don't have liquid as an option.

So, um, yeah, we're, we have really interesting solutions, uh, for, for liquid, which can take the absolute highest wattage, highest heat parts that exist. But we also have to solve the problem for error. And, uh, we've been doing a lot of work on figuring out how to adapt the server design so that we can take on 10, 650 watt power supplies for exam, uh, excuse me, uh, 650 watt GPUs, um, in a four U server, and still be able to cool them plus cool the, um, the SSDs and the CPUs and the memory that's in those servers, right?

You know, we focus on those GPUs, which is very understandable, right? 'cause they're really high, high power. But we still have the CPUs and memory and SSDs as well that have to get, oh, and as is my friends always remind me, you gotta always talk about those, uh, transceivers, right?

The, uh, the optical transceivers are also now power draws and they also demand, um, a lot of air cooling. So, uh, yeah, that's one of the things we focused on in gen 12 is ensuring we have really strong options for customers that are doing AI that, that need those high water GPUs and being able to continue to use their air cooling for those solutions. I, I remember going to Houston years ago and seeing some of the incredible stuff.

You guys were engineering your own fans, you guys were, you know, and obviously, you know, you've got a incredible yes, um, work done on baffles and heat sinks and so on. I remember seeing all that stuff in person, and then I remember opening up a server and, and seeing it right there in front of me and saying, yeah, this stuff from the lab. Here it is on the, on the server.

Yeah. We, we, you know, I, I got, we take tours of people through our facility here in Houston, and a lot of people are surprised to learn that we have wind tunnels, by the way I say wind tunnel. And immediately someone thinks that it's some big thing you can get a car or an airplane In our wind tunnels are very small.

They're designed to put a server in or they're designed to just put the fan assembly in. Um, but yeah, we design fans, we design shrouds, we design all of baffling. Um, Pete Sink is a really, uh, hot topic, no pun intended, uh, right now as we are working on, like I said, on, on trying to figure out how to do that.

Um, but you know, one thing we tell customers, and you know, it's just, it's really simple, is that, you know, losing density can often buy you a lot. Um, if you think of the exact same server load, say an 850 watt load, just add it up, you know, all the components in a one U server versus a two U, the one you will consume an extra hundred watts for its fans alone to cool that 850 watt load, if you put it into a two watts, a two U server, it only takes 17 watts of fan power. And, and then when Scott, when you sit that on underneath a desk in somebody's little manager's office, right?

Yeah. It changes the, so it changes it. So it doesn't seem like you have a jet engine running under your desk.

My office server under my desk was just fine. Thank you very much. No, no.

The acoustics are important. Yeah, we actually do care about that as well for those servers that are clearly designed for edge locations. Yeah.

Because you don't know where they're gonna end up, right? And you can't have it next to someone's head running like that. But, uh, but no, you know, we we're trying to help customers understand that if they don't need density, don't pay for it.

And what I mean is not, not paying me paying the power budget, so you go lose density, it's a good, it's a good thing and it'll really benefit you. But I, I think he's hit something very important there, Scott. You know, Stevens mentioned, do you need servers?

Well, yeah, we don't need, don't need commodity servers, Jack. We need engineered solutions. Something with a lot of, with a lot of thought behind it that solves problems that you don't even think about and don't want to deal with.

Heat, power, longevity. So, and then when you have an organization that does look at the whole equation and solves that for you without you, you know, stressing about it, that's a win. That's a win right there.

And Well, when you talk about engineering too, we're not just talking about the server itself. I mean, we're talking rack scale in these cases too. I mean, you know, in, in, again, I have seen HP's production floor, you guys are building whole racks for customers, right?

I mean, it's not just the, you know, it's not just servers that you're shipping. Yeah, for sure. And I mean, not only racks, but rows and, and depending on what the customer needs, right?

We, we have HPC systems, which are multiple rows and can consume, you know, uh, you know, mul thousands of servers and tens of thousands of GPUs. Uh, we can build everything from, from yes, one server all the way up. You guys are the number one supplier in H-P-H-P-C, right?

We are, and we are the top of all the big charts, uh, uh, in terms of the best and the, the best supercomputers built from Frontier to Elk, Capitan to, uh, as the most recent two. But, uh, yeah, the first, uh, some of the first and, and largest supercomputers ever built. Incredible.

Wow. Um, so I, I think that as we're kind of, uh, you know, getting here, uh, on, in the episode, uh, honestly, I think all of us are of, of agreement that servers still matter. I mean, that's why we're coming out here, right?

That's why we're flying to Houston. To to, to visit h hp HPE and learn a bit more about what they're doing. I almost said HPC, um, I guess we're visiting HPC too.

Um, learn about, uh, what, what you're doing in HPC and AI and Edge and so on. Um, you know, Jack, what are you looking for in a conversation with HPE about servers? Um, I'm looking to see how we, the conversation has moved from what used to be server consolidation to really now server expansion and what are all the different ways HP is meeting the customers, these new use cases.

And in particular, thinking about all these various use cases. How are we going to secure servers and the server environment? Um, I'm, I'm, you know, big on security, big on ai, and these are the things I think that are gonna drive computing in the near future.

And so what is HP doing to meet these customer requirements and the new workflows and new ways of thinking about the computing world? How about you VI know that you know the HPE product line very, very well, but, uh, what are you learning? What are you expecting to learn here?

Uh, you know, I, I wanna see some of the people behind the scenes. I wanna see their thought process, what, what drives them, what motivates them and what, and how they bring that to the HBO product line, because that, that ethos is so important, and that, that's part of that culture of HPE has always been engineering, and people don't get it. And I think they need to really leverage that to help people understand that HP does care about your, your, your compute requirements, care about your mission, your outcome, and this is why.

It's just, that's not a need to copy this, copy that, and we'll just tear apart. And it's, it's more than agnostic. It's purpose built.

So we, we need to help them understand it's purpose built so we can derive the outcomes that are expected and exceed their expectations in most cases. 'cause how sometimes how old is your, your, your little proline, Steve, it's probably like, and it's still chunking along, right? Oh, come on.

It's a Gen 10 man. It's not that, all. Right.

Right, right, Right. I've seen up times, amazing up times with HPE servers and, uh, it, it, like, again, it's the balance between software and hardware and how they're really engineered well. So, kudos And Scott, um, I guess, what's your message?

Uh, you know, so, so when this e when this episode airs, uh, it will be the day of the event. Um, we will be, uh, coming live on, on LinkedIn, on, uh, tech Strong TV and so on. We're gonna be posting these videos on YouTube.

Um, what do you want the audience to learn about what HPE is doing in the server market? Well, yeah, what I'm hoping is that, um, and get a sense of that, you know, we really, uh, we really do care about building the best, uh, server system possible. That we work hard to engineer it.

It's by design to be secured. It's by design to be optimized. It's by design to be automated so that it, uh, can fit in.

All the servers we build can fit all kinds of different locations, as well as, of course, core data centers and be the best, uh, solution we can build. I mean, our, our whole goal, we really do believe that we build the best servers in the world. And I hope, I hope that comes across Excellent.

I mean, I, uh, I think it's, uh, well, I, I've had very good experiences with HPE ProLiant, let me just say that. Um, uh, thank you very much, uh, for showing this, and I'm very glad that we are able to, to kind of give people a look inside, like one was saying inside the, the, the factory, inside the server factory. Talk to the people.

Learn from the people who are doing the work, who are engineering these next generation servers. I know we're on gen 12 now. Um, is it here, Scott, can I put you on the spot?

Is it gonna be a Gen 13 or are you gonna skip 13? Like, like a hotel elevator? That's, people bring that up all the time.

We absolutely are working on a next generation. What it, what it gets named is, uh, kind of not up to me in engineering, but, uh, we'll see what the marketing team comes up with. We'll see, we'll see Indeed.

Well, thank you very much for this. And, uh, those of you listening, uh, please do, uh, check out the Tech Field Day website. You can see a lot more detail about these servers.

Basically all of the things that we've been talking about here today, including a very, very sweet little edge server, uh, the DL 1 45 that was, uh, recently released. Um, check out on YouTube as well. If you've missed these things, you, you just Google, uh, tech Field, A HPE, ProLiant, you'll find the videos there.

Um, and I think you're gonna learn something. Even even VM who knows more about this than me. I, I is for sure is gonna learn something.

I guarantee it. So thank you for listening to this episode of the Tech Field, a podcast. If you enjoyed the discussion, please do subscribe.

Uh, you can find us on YouTube. You can find us in your favorite podcast application. Uh, please also give us a rating or review.

We'd love to hear from you. This podcast is brought to you by Tech Field Day Home from IT experts from across the enterprise, which is part of the FUTURUM Group, and was recorded in association with our Special Tech Field Day exclusive event that we're doing, which was sponsored by HPE for upcoming episodes. com/podcast or find us on Textron tv, or as I said, in your favorite podcast application.

We'd also have, uh, accounts on X Twitter, uh, uh, blue sky and, and Mastodon. So we'd love to hear from you from there. Thanks for listening, and we will see you next week on the Tech Field Day podcast.