Clarence Reynolds, TelecomTV (00:03):
We are at the last day of Mobile World Congress 2025 here in Barcelona. And it has been a fantastic show and it takes a lot of people to put this show on. And we have two of them that are here who are here to tell us all about what's happening here at Mobile World Congress. And Andy Bryant is part of the worldwide marketing team at Hewlett Packard Enterprise. And Adrian Kasbergen is senior content director also at HPE. Thank you for joining me today. And I want to say this has been one of my favorite M wcs.

Andy Bryant, HPE (00:35):
It has. It's been very busy this year. We've had a lot of people coming around the booth. It's been buzz the whole time.

Clarence Reynolds, TelecomTV (00:41):
And there's this thing called AI, I guess that everybody's talking about.

Andy Bryant, HPE (00:46):
Yeah. AI has been a hot topic during the show. This is possibly one of the biggest ones.

Adrian Kasbergen, HPE (00:51):
Yeah. AI has been a buzz, not just during the show, but for the last year, two years. And increasingly and when we designed the content for this booth, we knew that we needed to give prime spot for ai. And so what we're showing here, AI is not just generative AI do a chat bot or what have you. We show how we use AI in network management, in system management, how we make AI very easily deployable by selling a complete rack that is completely provisioned hardware software to have AI up and running at a customer in a matter of hours to days rather than months to years.

Clarence Reynolds, TelecomTV (01:36):
And one of the things I think that HPE does as well or better than anyone else, is partnerships. And one of the things that when we got to a point in 5G that you start to notice is that we couldn't go any further without playing together and without working together and without really these collaborations that are a part of it. Has that really been something that you've seen ramp up over the past couple of years?

Andy Bryant, HPE (02:02):
Yeah, I think I was listening to one of the analysts earlier talking about the wrap up of the show, and they mentioned that partnerships has got more and more popular for this event. And we're here with our partners with a MD, with Intel, with Oxia,

Clarence Reynolds, TelecomTV (02:14):
Nokia,

Andy Bryant, HPE (02:15):
With Nvidia, with Nokia. We've got solutions integrating all of their components.

Clarence Reynolds, TelecomTV (02:20):
So tell us about the booth. Has this been about normal for traffic? Has it been sort of more buzz because there are so many partnerships and AI is such a big thing?

Andy Bryant, HPE (02:33):
Yeah, we've had a lot of people interested in the big, shiny, the Cray we have at the back of the booth, but also a lot of interesting conversations from customers that are looking to get deeper into the technology. So last year I think AI was kind of, it was a buzz, but it wasn't really what can you do with its business. It was kicking the tires. This year we're getting more deeper conversations about how you can use AI to dramatically lower costs through operations, through AIOps for the telco network, how you can use it in the RAN with our AI Ran Alliance work. It's getting further and further embedded into the technology, I think.

Clarence Reynolds, TelecomTV (03:07):
And one of the things I'm seeing is how well designed the booth is because you walk around and you get a sense of the variety of things that HP is working on or has already worked on and these platforms that are emerging. Is that part of the thing to draw people in to say, listen, this is a working booth, we've got things here that actually can help your business grow, help you save money more efficiently, work more efficiently?

Adrian Kasbergen, HPE (03:34):
Yeah, so from the functionality for the customer point of view, you are spot on. You saw that well, but also especially this year, we really looked at the possible traffic patterns, making sure that it was always easy for a visitor, for a customer to walk from one demo to another. And as Andy said, and I've been observing that all week compared to previous years and compared to other shows, there were many, many longer conversations, deep conversations where I was looking at a certain demonstration and it's like they've been at it for 20, 25 minutes, which normally is an exception. And here it became the rule. And the other thing that then we were just talking about that before the interview, Thursday last day. Typically that's a slow day. And even right now there's still a lot of people on our booth, but until 12 o'clock to Booth was hard to maneuver through because of the amount of traffic that we got.

Clarence Reynolds, TelecomTV (04:37):
I have had a chance to see a lot of what's happening at the booth, but a lot of people haven't seen it from that perspective and through your eyes. So I want to take a walk through and just to see some highlights.

Adrian Kasbergen, HPE (04:50):
Sure,

Clarence Reynolds, TelecomTV (04:51):
Absolutely.

Adrian Kasbergen, HPE (04:51):
Let's do that. Let's have a look at multi-vendor network management where we combine the power of Aruba Central with the capabilities of Ops ramp. They use a common data lake in GreenLake that gives the opportunity for our customers to manage the Aruba networking elements as well as all of their IT infrastructure, be it compute, networking and storage into a single pane of glass.

Andy Bryant, HPE (05:37):
We announced this week the industry's first enterprise networking switch with precision timing integrated particularly useful for AI workloads, open RAN or C ran. And also for providing the precision timing you need to hand off between access points in a large public venue if you're running private 5G. So rather than having the precision timing feed coming in from satellites, this one actually has it embedded in the switch. Now let's roll around the corner. Aruba networking connected Edge AI and private 5G solution. So here we've taken our private 5G stack. We've packaged it together onto a ProLiant deal, 3 25 server to create a turnkey solution. The objective here is to make private 5G as simple to deploy as wifi. So make it available to more enterprises with more scale.

Adrian Kasbergen, HPE (06:25):
Okay, and let's have a look at what we're doing with FC Barcelona. As you may or may not know, FC Barcelona is rebuilding their entire stadium in place. And for that purpose, they have vacated their old stadium for a period of time. They're just back in it. And now we're completing the inside facilities of that stadium. From a networking perspective, we're enabling them to provide their visitors maximum performance, wifi and 5G capabilities as well as their back office where we're connecting their signage, counting people, occupancy control, people tracking, all using AI to interpret the data that is coming in and help them make the best decisions.

Andy Bryant, HPE (07:15):
So we do a lot of our services for enterprises for private cloud. We also have governments and service providers that want to build their own solutions to offer within their sovereign areas for sovereign private cloud, sovereign AI for AI factories. And we're providing the capabilities in partnership with NVIDIA to build out that infrastructure to allow them to offer those services to their customers from within regions such as the European region.

Adrian Kasbergen, HPE (07:40):
So what we're showing here is part of our high performance computing line, HPC, and in this case, these are platforms that are optimized for AI by using very large counts of GPU capabilities. In the XD six 80 platform that we're showing here, we've taken out the compute tray and that compute tray has eight intel gdi three AI processors to provide a performance that wasn't seen in this type of platform before. And for people who prefer the NVIDIA platform, we have a similar solution, but based on the NVIDIA H 200 processor.

Andy Bryant, HPE (08:24):
Okay, so here we're talking about our disconnected cloud for sovereign environments. So we've built a private cloud solution for enterprise customers, but we found that some customers, particularly in the defense space, don't like the private cloud to be connected to the public cloud. So we've put the management solutions for that into the system itself so that we can run a completely disconnected sovereign or private cloud solution.

Adrian Kasbergen, HPE (08:47):
Let's have a quick look at multi-vendor network management where we combine the power of Aruba Central with ops ramp IT management to give our customers through a common GreenLake data lake, the capability to manage through one pane of glass, both their IT equipment and their networking equipment. So what we are showing here is part of a supercomputer straight off of the manufacturing line, and this is only a small part of the entire supercomputer. We see here on the left side a distribution rack for the liquid cooling that then goes over the top to all of the compute racks. We can have in a situation like this, up to four compute racks. And then we have, depending on the customer's needs, 5, 10, 20, 30 of those rows of compute racks. Each compute racks consist of eight modules with eight servers. So 64 servers in one cabinet, which leads to 81,920 CPUs, A CPU course or 448 GPU in a single cabinet. That creates a lot of compute power. The liquid cooling provides more efficient cooling than would ever be possible with air cooling and also provides better sustainability. Where we see that some of our customers hook up the liquid cooling system through heat exchangers to warm, for instance, community pools or apartment buildings.

Andy Bryant, HPE (10:28):
And it's our solutions for telecom's operators. Here we're talking about our HPE ProLiant deal one 10 Gen 11 and the Gen 12, which is just around the corner, including the Intel Zion six latest generation of processor with Granite Rapids D providing solutions specifically for ram. And here we're also showcasing the Nokia Cloud RAN unit adjacent to Telco. We have service providers, the DL three 40 optimized for the service provider market and then inter compute optimized for telco core. So in this case, we're showcasing the A MD epic processor optimized for high performance, high reliability and low power. So particularly for telco core workloads still popular on the last date. And then last but not least, in the telco space, we have the AI automation for telco core and ran, this is where we're showing the move towards autonomous networking in the telco space from assurance through orchestration and fulfillment.

Adrian Kasbergen, HPE (11:26):
So what we're showing here is a solution that we're very proud of, which is the HPE private Cloud ai, in this case for telco ops, but there is a regular business solution as well. We've developed this together with Nvidia and the key of the solution is that it has all the hardware, all the software, all the utilities, all of the tools on board to get private cloud AI up and running in no time at all. And not just that we sell it ready to run, but also in different sizes for different customers. So if you're a small customer, you start with the small system. If you're a medium sized customer, you get the medium large or even extra large. And so this is very popular both for customers in their end environment. They can get a AI solution up and running in a matter of hours, days instead of months to years.

Andy Bryant, HPE (12:26):
So we just covered our solutions for the infrastructure for telco here. We're talking about our telco software team with OSS solutions, driving towards autonomous networking. So we have products in this space all the way from assurance through orchestration that help to automate the network applying AI to the telco network. But that's not the only AI we have in the telco. We also have, we are a member of the AI RAN Alliance and over here we can see the Proli DL 3 84, gen 12, including the Nvidia GH 200 Brace Hopper Compute. In this place, we're showcasing how telcos can mix or separate out processing for RAN and processing for consumer workloads that might be running on the same compute out at the C RAN site.

Adrian Kasbergen, HPE (13:15):
So this is for me, really an exciting part of our show floor here. Hope we can show this at many of our exhibits showing the space borne computer. Since 2017, we've had computers in the International Space Station, starting with the space borne one, just running benchmarks. Now, the space borne two that is actually providing cloud services to experiments on board of the International Space Station. The chart behind me, I'm not going to go through all the detail, but it shows how much time is saved on the internet International Space Station in physical time and in compute time by having these supercomputers on board. So I hope you enjoyed our tour over the show floor as much as thousands of our visitors did, and that you will come back to us next year.

Andy Bryant, HPE (14:10):
Yep. We'll see you in 2026. Thank you.