Scott Meltzer, Comedy Industries (00:02):
Howdy everybody. We're going to get the next presentation started for you right here in the Hewlett-Packard Enterprise Theater in just two minutes, two minutes till showtime. Come on in, grab a seat. In this presentation, we're going to talk to you about the HPE Spaceborne computer, talk to you about our partners at Kioksia, and we are going to give away free stuff. Let me show you the free stuff we're giving away. Look at this right here.

Katrine Spang-Hanssen, Comedy Industries (00:24):
Oh, look at this. Now, these are our pair of high quality Bluetooth headphones, and this is our final presentation here at Mobile World Congress. So if you would like to win them, come on in, grab a seat. Currently, your chances of winning are quite high because we have how many people here? One, two, three.

Scott Meltzer, Comedy Industries (00:43):
Four, five, six, seven, seven. Four, five, six, seven. Seven people.

Katrine Spang-Hanssen, Comedy Industries (00:45):
So come on in, grab a seat. This is your best chance to win because it's the last chance.

Scott Meltzer, Comedy Industries (00:50):
Exactly. So we'll get started here in just 50 seconds. As Katrina mentioned, your odds of winning the Grand Prize are one in one, two, three, four, five, six, one and six. Those are good odds, but they're not good enough. So before anyone else gets here in our last 30 seconds before we start, we're going to double your odds. Instead of one in seven now, one in seven, your actual odds will increase to two in 14. Two and 14. That is twice as good as one in seven. Twice as good. Now sure, some of you are saying I'm an idiot because I can't do math and it's still the same odds. But you're not thinking it through. As you can see, there's a bunch of people standing in the back now. The seats are filled. So every other person who comes over, that means your odds go down.

(01:29):
But don't worry because you can increase your odds and here is how. We're going to start with a little quiz. If you answer a question correctly in the quiz, you will earn an extra ticket for the drawing. And the more tickets you get, the better your chances to win the headphones. Does that sound good? Yes is the correct answer. Some of the questions are easier than others. Oh, if you laugh at the dumb jokes, you'll get an extra ticket for the

Katrine Spang-Hanssen, Comedy Industries (01:50):
Phone. Yes. Oh, if we see you pay extra close attention like that right there, like nodding along as if I've said something really profound. Do you think that'll get you a ticket? Yes, it will.

Scott Meltzer, Comedy Industries (01:59):
If you take out your phone and you take a picture of one of our slides, you think that's going to get you a ticket? There's only one way to find out. Oh, only

Katrine Spang-Hanssen, Comedy Industries (02:06):
One way

Scott Meltzer, Comedy Industries (02:07):
To find out. The first one gets the ticket. Let me see that. That was a selfie. You have to turn the camera around. Let me see

Katrine Spang-Hanssen, Comedy Industries (02:12):
Your picture right there. Okay. Oh,

Scott Meltzer, Comedy Industries (02:14):
That's a good picture of the slide. I like that.

Katrine Spang-Hanssen, Comedy Industries (02:16):
Okay. That's a picture of me.

Scott Meltzer, Comedy Industries (02:17):
I'm okay with that too. I'm fine with that.

Katrine Spang-Hanssen, Comedy Industries (02:19):
That's all good.

Scott Meltzer, Comedy Industries (02:20):
Now, when you look at this slide, it may be a little confusing to some of you. It is very wordy. I know this. And English might not be everyone's first language. So actually for another ticket, who here English is your second best language? Second. Okay? Okay. I see you. I see him over here. All right, very good. Okay. Who here English is your third language? Third, best, third. Okay. Okay third. Here in the front. Third. Good third. Who here? English is your fourth best language. Fourth, fourth. One guy over here whose English is getting worse every time he raises his hand. That's amazing. But I am impressed. Four different languages. I find that very impressive.

Katrine Spang-Hanssen, Comedy Industries (02:58):
Well, actually, as I keep telling my American colleague here, it's not that big of a deal here in Europe. I, for example, I'm from Denmark, so I picked up Danish, Swedish, Norwegian, and English just by existing. In fact, watch this. Who here speaks only one language? Oh, where in the UK are you from?

Scott Meltzer, Comedy Industries (03:20):
Exactly. Exactly. But the point is, no matter what language you speak,

Katrine Spang-Hanssen, Comedy Industries (03:25):
No matter what country you're from.

Scott Meltzer, Comedy Industries (03:27):
HPE has the secure AI native infrastructure you need.

Katrine Spang-Hanssen, Comedy Industries (03:31):
Infrastructure that's purpose built with AI and for AI.

Scott Meltzer, Comedy Industries (03:36):
Or as we like to say here in Espana.

Katrine Spang-Hanssen, Comedy Industries (03:45):
I know we have a lot of Spanish

Scott Meltzer, Comedy Industries (03:47):
Speakers here

Katrine Spang-Hanssen, Comedy Industries (03:47):
Today. How do we want to grade his accent?

Scott Meltzer, Comedy Industries (03:50):
On a scale of perfecto to American. Perfecto. See, that's what we call sarcasmo. Sarcalsmo. Very good. But this brings us to our first word search.

Katrine Spang-Hanssen, Comedy Industries (04:05):
So for a ticket, who here works for either a telco or a service provider? All right.

Scott Meltzer, Comedy Industries (04:12):
That's a few of

Katrine Spang-Hanssen, Comedy Industries (04:13):
You.

Scott Meltzer, Comedy Industries (04:13):
Okay, good. That perfect. And that is very useful because that's exactly what our next slide is all about. Modernizing telco networks. Oh, and for ticket, who here can find the word modernize, hidden in our puzzle? Nicely done, sir, right there. Modernized. Excellent. Pointed right at it. Or as we say here in Aspanya, modernity.

Katrine Spang-Hanssen, Comedy Industries (04:37):
Modernity. And so if you want to modernize your telco network, who should you talk to? Yeah. Okay. I think we didn't make ourselves clear.

Scott Meltzer, Comedy Industries (04:49):
If you want to modernize your telco network, who should you talk to? Much better. Only apologize to the people in the back if the letters were too small for you to see. That wasn't fair to you. That's not fair to all the people in that. How could you possibly get it all the way back here? And if you do want to modernize your telco network, let me tell you, we have got the best solutions for you, including AI native routing to provide high performance at scale.

Katrine Spang-Hanssen, Comedy Industries (05:15):
We deliver secure AI at the edge for Denser ORAN and VRAN deployments and AI RAN acceleration.

Scott Meltzer, Comedy Industries (05:22):
And we help you simplify your telco deployment and operations.

Katrine Spang-Hanssen, Comedy Industries (05:26):
And in fact, simplify, that is a word we love here at HP because we're all about simplifying your network. So can anyone find the word simplify on false in French up here? If you're not a French speaker, the word we're looking for is simplifier.

Scott Meltzer, Comedy Industries (05:42):
And I want to point out the words can go forward and backward, but they also can go up and down. Can anybody find- Can anyone find it? Just point to

Katrine Spang-Hanssen, Comedy Industries (05:49):
It. Simplify

Scott Meltzer, Comedy Industries (05:50):
It. Can anyone find it? Jus point. Nicely done. Thank you. Thank you nicely done. And of course, because we simplify operations, can anyone find operations up here in Italian? In

Katrine Spang-Hanssen, Comedy Industries (06:03):
Italo.

Scott Meltzer, Comedy Industries (06:04):
Italiano.

Katrine Spang-Hanssen, Comedy Industries (06:06):
Nicely

Scott Meltzer, Comedy Industries (06:06):
Done. Right next to it. Right next to it. Beautiful. Operate Yoni. But not only do we simplify operations, we also accelerate your data center AI workloads.

Katrine Spang-Hanssen, Comedy Industries (06:15):
Oh, if you need to boost AI performance for those applications, well, you need cutting edge networking solutions that are designed specifically for AI. Excuse me. Excuse me. I have a question. Oh yes. Over here. Yes, yes.

Scott Meltzer, Comedy Industries (06:31):
I

Katrine Spang-Hanssen, Comedy Industries (06:31):
Want to know what about your multi-vendor AI ops solutions.

Scott Meltzer, Comedy Industries (06:35):
Thank you so much for asking.

Katrine Spang-Hanssen, Comedy Industries (06:36):
Oh, thank you. You're talking about the AI ops solutions that save you time and money without locking you into a specific vendor or product, those?

Scott Meltzer, Comedy Industries (06:44):
We, we. I'm from France. I did not know that when I started talking. Yes. What word would use for those?

Katrine Spang-Hanssen, Comedy Industries (06:50):
Thank

Scott Meltzer, Comedy Industries (06:50):
You.

Katrine Spang-Hanssen, Comedy Industries (06:51):
Thank you for asking that very important question. Because I'm from Denmark, I would use the word which means excellent Danish, but I would also have settled for the best. But since we have so many Spanish speakers here today, now can anyone find the best in Spanish up here? If you're not a Spanish speaker, the word we're looking for is. Can anyone find Lomehoor up here?

Scott Meltzer, Comedy Industries (07:18):
Yes, sir. Right. Beautiful. Right there in the center. Nicely done, sir. Very quick. Yeah, good. And if you want Lomecor Soluciones Parallel Centra the data. I don't speak Spanish. Yes. Oh yes. Excellent. Excellent. That's good. That's good. Thank you. Yes. Thank you. If you want the best solutions to eliminate data silos so you can train your AI models on all your data without worrying about your competitors also training on that data. Well, one company has the solutions you need and that one company is- One company is? HPE. Very good. Absolutely. They've got that IP. Absolutely. Yes. It is HPP.

Katrine Spang-Hanssen, Comedy Industries (07:59):
And in fact, this solution we're talking about here is HPE private cloud for AI. This is the turnkey AI solution you need.

Scott Meltzer, Comedy Industries (08:08):
Right. Now, turnkey means that everything comes pre-installed and pre-integrated. So you can get those AI solutions up and running in minutes instead of like, I don't know, like months for the other guys. All

Katrine Spang-Hanssen, Comedy Industries (08:18):
Right. So for a ticket.

Scott Meltzer, Comedy Industries (08:20):
Oh.

Katrine Spang-Hanssen, Comedy Industries (08:20):
What's there not to love about that? Nothing. Nothing exactly. Exactly. And to all the people who said nothing in the back, you get a ticket because-

Scott Meltzer, Comedy Industries (08:29):
Oh, you get a ticket for saying nothing very good. And all around that five, you can see some great demos of how HPE can help you transform enterprise experiences with secure AI networking platform and innovative networking solutions. Because

Katrine Spang-Hanssen, Comedy Industries (08:45):
When it comes to delivering the most comprehensive client to cloud portfolio,

Scott Meltzer, Comedy Industries (08:51):
All unified under a single platform,

Katrine Spang-Hanssen, Comedy Industries (08:54):
HPE is the leader in so many different areas, including, and here you can just read it off the screen. Including?

Scott Meltzer, Comedy Industries (09:02):
Yes.

Katrine Spang-Hanssen, Comedy Industries (09:02):
What else?

Scott Meltzer, Comedy Industries (09:03):
And? Wifi. Absolutely.

Katrine Spang-Hanssen, Comedy Industries (09:06):
And?

Scott Meltzer, Comedy Industries (09:07):
Profit 5G, of course. And what else? When? Yes. Yeah. All these are important, but what do you want most of all? What brought you over to our booth today? No, no, no, no. Fewer troubled tickets. Fewer troubled tickets.

Katrine Spang-Hanssen, Comedy Industries (09:23):
And here at HPE, we have eliminated 85% of truck rolls and 90% of trouble tickets for so many of our customers with our cloud native and AI native self-driving operations. But none of

Scott Meltzer, Comedy Industries (09:37):
This is possible without the right security. And oh, where do you find the right security? No, no. We're in the puzzle. We're in the puzzle. We're looking for the word security in Catalan. In Qatalan. What is security in Qatalan? Security that. Very good. Yes. And who can find it up here in our puzzle? Secure that. Nicely done right over there. The gentleman right over there. Beautiful.

Katrine Spang-Hanssen, Comedy Industries (10:02):
He's figured out you could just point at the

Scott Meltzer, Comedy Industries (10:04):
Screen. Just point at the screen. Just point at the screen.

Katrine Spang-Hanssen, Comedy Industries (10:06):
And here at HPE, we give you secure infrastructure at scale.

Scott Meltzer, Comedy Industries (10:11):
With scale out firewalls to protect your data center traffic.

Katrine Spang-Hanssen, Comedy Industries (10:15):
And carrier grade security to maximize uptime with customizable and scalable solutions. But

Scott Meltzer, Comedy Industries (10:22):
Only one company can provide this kind of secure AI native infrastructure that is purpose built with AI and for AI.

Katrine Spang-Hanssen, Comedy Industries (10:30):
Only one company gives you integral security to help you enforce corporate policies across users, devices, and data for everyone, everywhere.

Scott Meltzer, Comedy Industries (10:42):
Only one company provides truly open solutions that can eliminate all the boundaries. And that one company is- HBE. Oh, thank you so much. Thank you so much. We are going to bring up our featured speakers right now. They're going to talk to us about super computing in space. And then right when they're done, we're going to do our drawing. Now, I want to make sure that everyone here has a chance to be in the drawing. So if there's anyone here, if you got here late and you did not get a ticket for the drawing yet, raise your hand if you did not get a ticket yet, right over here. Very good.

Katrine Spang-Hanssen, Comedy Industries (11:21):
Or

Scott Meltzer, Comedy Industries (11:21):
If

Katrine Spang-Hanssen, Comedy Industries (11:21):
You're willing to lie for a ticket.

Scott Meltzer, Comedy Industries (11:24):
Ah, very good. Even better. Very good. This man is going to be an executive someday. That's excellent. He has all the important features.

Katrine Spang-Hanssen, Comedy Industries (11:31):
All right. So folks, we are just moments away from introducing our featured speaker. We have absolutely saved the best for last year at Mobile World Congress. Folks, you are in for tree. We have a great partner here at HPE called Kioksia. And along with the HPE, they have helped send computers into space. And that's what we're going to talk about how our space program has enhanced computing out there in the universe. So it's super exciting, a super fun presentation. And so therefore, the people who clap the most for our featured speakers, they get the rest of our tickets. So please welcome to the stage. And of course, right after they're done, we raffle off the headphones. Please welcome to the stage Igor Potter and Norm Follett.

Norm Follett, HPE (12:27):
All right. Thank you very much, ladies and gentlemen. So we're going to have a little bit of a two-part discussion. My good friend here, Igor Potter, will explain who exactly Kioki is. And then we're going to take a little trip across the universe. So first, I'm going to pass the baton to Igor and he's going to kick us off.

Igor Pater, KIOXIA, HPE (12:45):
Yeah. Thank you, Norm, very much. So before I begin, I just would like to have a quick show of hands. Who here knows what Kyoxia is? Excellent.

Norm Follett, HPE (12:55):
Excellent.

Igor Pater, KIOXIA, HPE (12:55):
Okay. So we have some work to do. All right. So Kioxia is a leading worldwide SSD and flash memory supplier. You might be wondering how does a company seemingly out of nowhere become a leading supplier in such a mature industry? And the other is simple. It's not out of nowhere. It's not a new company. Prior to 2019, you might have heard of a company called Toshiba Memory, part of the Toshiba Conglomerate. And yeah, they spin out from the conglomerate, they formed their own company, and in 2019, they changed their name to Kyokia. So although the brand might be new, the people behind it are the same people that have invented flash memory almost 40 years ago. I'll talk about the factories in a second, but what I would like to highlight with this slide still is that Kyoxia remains vertically integrated. This means that they produce their own flash in their own memories.

(13:47):
They design their own controllers and they write their own firmware that later can be customized for OEMs like HPE, of course. Now a bit of geography. Keyoxia is headquartered in Tokyo, Japan. Japan is also where they are producing all of their worldwide flash supply with UK plan that you can see at the bottom over there, being responsible for 30% of worldwide flash memory supply. That is not 30% of Keyoxa supply, this is 30% of worldwide supply, making it probably one of the most important factories in the world right now. And quickly about the solutions. So Kioxa has two main business units, SSDs, both for the data centers and for the customer clients, so for your PCs or your laptops, and then memories. So think about your smartphone, smart fridge, smart car, whatever is using flash memory, whatever is smart, there's a high chance it is going to be using Keyoxa solutions.

(14:46):
And with this being said, I want to cover everyone's favorite topic, AI. So if you are with us here last year, you might have heard us talking about how AI is affecting the SSD supply and demand. If you are trying to buy a drive this year, you know exactly how it affects it, but that's not what I'm going to cover today. Today, I want to go over the AI data flow and explore how the SSD plays crucial role at every stage of this data flow. And even more importantly, why one solution does not fit all, and you need the right SSD for the right AR workload to maximize your effects of the whole infrastructure that you have. So let's start where the data first enters your systems, meaning during the data ingestion. As we know, these deep learning models, they can have huge data sets.

(15:38):
They require a highly scalable and cost optimized solution. Thankfully, Qokta has you covered and we have a solution right for you. So built on industries first, two terabyte BIX8 QLC based, that's a mouthful. If you want to know more, come to my book later. Technology, LC9 series from Keyoxia is their highest density solution starting at 30 terabytes and going all the way to hundreds of terabytes in OneDrive. In fact, recently you might have seen a photo or a video of Linus tech tips holding four drives in his hands saying that he has one petabyte of data in his hand. These are actually Keyoxa drives. Then we go to AI training. So during this phase, the last thing you want your GPUs to be doing is idling and waiting for the data to flow in. So you need a solution, SSD that will provide you with low latency, and high bandwidth in both reads and writes.

(16:35):
And you might have guessed it. Keyoxec has a solution for you. So depending on your server platform needs, this will be either CD9P for the data center customers, or in case of hyperscalers, this will be the XD8 series. Now we are going to the inferencing and during the inferencing phase, you need to start servicing multi-tenant and multi-model environments. So you need a solution that will be still low latency, but read optimized. And yes, you might have guessed it again. The solution from Qyoxia will be ... Oh, sorry, not this drive. It'll be this drive, the CD9P, but in the read optimized version. And last but not least, the RAG. During this phase, the vector databases for your RAG solutions can multiply the size of your dataset 10 times or even more. So you need a best performing solution to serve this need, and this will blow your mind.

(17:34):
Keyoka has a solution for you as well. Based on a flagship CM9 series, these are the best in class performing and the lowest latency drives you can get on the market right now. But this is what we have today. Now let's quickly have a quick glimpse into the future. At the moment, the best performing drives on the market can reach around three million iOPS. This is huge amount, but we know that for certain workloads, for certain customers, this is not good enough. So we are collaborating with our customers and our ecosystem collaborators in order to develop the first 10 million iOPS drive, hopefully available later this year. This will be based on the Excel flash second generation technology from Keyoxia as well as the development of PCI gen six standard. So this is the first step. Then much more ambitious goal is reaching 100 million iOPS from OneDrive.

(18:30):
We are hoping to achieve it by 2027 with, again, development in the technology of Excel/generation free, as well as the PCI Gen seven. And with this being said, I would like to hand it over to Norm. Thank you

Norm Follett, HPE (18:43):
Very much. All right. Thank you very much, Igor. Really well done.

(18:47):
So just to emphasize a point there, if we took a survey right now, 80% of you with your cell phones would have a Kioksia memory chip inside of it. So that includes Apple and the other vendors out there, so there's strong relationships. So now you know all about Kioksia. So what do we do with all that kind of memory and capability? Well, we put it on a rocket and we shoot it off to space and we make memories in space. So Hewlett-Packard Enterprise and the legacy company that was founded by Bill and Dave, Hewlett-Packard has actually been in the space race since the very beginning, since even pre-Mercury. So our founders, again, Bill and Dave, actually as electrical engineers personally worked on the problem of trying to communicate with rockets that go over the horizon. So that's where we were in the late 50s and early 60s, just trying to figure out how to talk to the rockets, have telemetry, share information.

(19:44):
Now that relationship with NASA continued and it actually went all the way through into the Apollo program. So when Neil Armstrong and Buzz Aldrin walked the surface of the moon in 1969, they actually had pins and diode switches built by Hewlett-Packard in their suits walking the surface of the moon. So HP, HPE has been to the moon already and before. They also did a whole variety of scientific testing and benchlining, understanding the health and human factors when we started sending human beings up into space as well. So also part of that. That relationship continued to evolve. How many of you guys have or have used a scientific calculator in your life? Engineers. Anybody have one in their pocket right now? Somebody must, right? Or you have an emulator. So the scientific calculator invented by HP, that program was accelerated and focused to facilitate the Soyuz mission, Soyuz and Apollo docking, right?

(20:48):
They couldn't speak the same languages. The navigation systems couldn't talk, but they could talk math. Now, how many of you remember in the film, Apollo 13 when they're doing all the calculus by hand? Remember that whole episode, all the human computers figuring it out? Another reason Bill and Dave focused on the scientific calculator to make sure that that type of scenario, that we were prepared for anything in the future. So that led us into the Space Shuttle program. So this is Sally Ride, the first American woman in space. And if you look closely, she's floating around in the space shuttle with a whole bunch of HP calculators floating around her. So the calculator, the scientific calculator was backup equipment on every space shuttle mission, every space shuttle mission. That of course, we're not done. That led us into something that we did, and we call that the Spaceborne computer.

(21:42):
So the Spaceborne computer first went up in 2017. HP has the most powerful computer. HPE has the most powerful computer to ever go to space. We've had three versions. Spacebourne one was a prototype, and meaning NASA asked us to see if we could take a commercial off-the-shelf system, if it could survive. It had to be commercial, something available on the market, if it could actually survive the shake rattle roll of launch, if an astronaut crew that was not IT experts could deal with it, and if it could be integrated into a spaceship. Now, in this case, that was a spaceship that was designed 30 years prior, and if it could be performant for a year. And the answer was, yes, yes, yes, we can do that. And it was so successful after the first year, and we were doing just ... It's called HPC benchmark testing.

(22:37):
We had a mirrored system in Chippewa Falls, our HPC factory, and we're just hammering this thing just to make sure and see what kind of performance it had in testing to see if there was any delta, any drift, any characterization, anything different happening at 254 miles above the planet's surface. And it turned out there wasn't. It performed exceptionally well. So then the light kind of went off for NASA and they said, "Well, while you're up there, could you actually do a little bit of work?" So we started to pick up operational tasks. Now, the Spaceborne computer has nothing to do with the avionics of the station. We are a scientific platform independent from the actual operations of the ship, but we are a scientific platform for the international scientific community. And our partners in that effort is the ISS National Lab. So on the space station, there are different segments and different locations each owned by different countries, and they each have different purposes.

(23:31):
The first time we went up, we were in the American scientific platform module, and that's called Destiny. Now we're actually in the Columbus module, which is the ESA module. There are two of these, two of these racks. It's a redundant system, and they're set up in the ceiling of the Columbus module, and they're liquid and air cooled. So we're actually plugged into the plumbing of the ISS, circulating out heat. Pretty cool. They do a lot with water up there, like things that they just do a lot with water and how they work that out. SpacePorn Computer launched in 2017, SpacePort Computer two launched in 21. Now, this was really interesting. So first, this was on a pure SpaceX mission. This was on a pure Northrop Grumman mission, and that spacecraft is called Cignas. You hear a lot about SpaceX, but there's another supplier, and that's Northrop Grumman that's also servicing the ISS, and that spaceship is called Cignas.

(24:27):
And then we did a 2.5 retrofit and an upgrade with our partners, Kioksia, and we brought the most memory ever brought up to space, a phenomenal number. And then that has led us into SpaceBorn Lunar. And we're going to talk about that in just a moment. So as the edge goes, so goes Hewlett-Packard Enterprise. In 2027, we believe we're going to be part of a progression of exploration on the surface of the moon. Here's a little bit of a deep dive on a Spacebourne computer 2.5, 120 terabytes of Kioxia storage onboard. Now, one of the things when I said the light went off, NASA and all of our partners up there believed that they could get away, so to speak, with doing all the computational work using Earth-based systems. But what we showed is that you can actually be far more performant, far more efficient, and far more ... In the moment, if you actually have computational capability on board.

(25:36):
You might not know, but the space station actually endures a tremendous amount of loss of signals. So you'll hear it in space.LOS. Literally, they communicate with the satellites, and when they don't have line of sight or connections with the satellites, they're actually dark. And that happens far more often than you would think. So what does that mean? It also means you can't transmit data back and forth. So it's a narrow pipe. So this is the classical edge computational use case. Do the computational work at the point of data acquisition. Send back answers. Don't send back gigabytes of data to be analyzed. Now, the ROI of that, your phone bill. Think about the expense of communication. All moving all that data is extremely expensive. You can do the work right there in real time, send back kilobytes, don't send back gigabytes, and then you have the information on hand.

(26:32):
Now, we've done over 50 plus experiments as part of our relationship, and that number's growing constantly with the international scientific community. In fact, we actually have a process where if you have a good idea, we'll listen to you. You can submit it. There's even a link on our website. Submit your idea. Let's have a discussion about what you think you can do and what you want to do on that platform. One such experiment that we did is something that we call the glove experiment. So astronauts have what they call an EVA season on the ISS, and that's spacewalks. That's what that is. Now, there's a time of year where it makes more sense for them to go outside based on radiation and solar flares. They can go out really anytime, but it's just a better practice for them to go out at certain times of the year when we're in a certain place with respect to the earth and going around the rotation of the sun.

(27:24):
So what happens is those gloves wear out. The gloves that the astronauts are wearing and they're out there working in their wrench. And it's just like you and your garden.You're working with that ho after a while and that thing starts to wear out. Now, imagine if your glove tears and you're outside of the spaceship, very bad, right? Very, very bad. So what they do, and this was really surprised me, is when you come into the ISS and you take off your equipment, they take thousands of pictures, right? They just take thousands of pictures of those gloves. Then they send all those JPEGs down to a room in Houston at Johnson Space Center and five guys sit in a room and a couple of ladies too, and they all sit there and they just look at the gloves and they go, "Hmm." And they go, look at the next picture, they look at the next picture.

(28:09):
And then they communicate up with the astronauts. "Hey, check pair number 17 index finger left side. Does it look worn?" And everyone's looking at it, "Well, I don't know. " And then they decide whether or not that glove is fit to go out. That process takes weeks. It literally, that back and forth process takes weeks. So guess what? They bring a lot of gloves because they're in the EVA season, so they throw those ones off to the side. They wait till they're re-certified to come back into service and then they go back out and they go back out with another set. So what we ended up doing, because first we were just communicating the JPEGs and moving the data around and all of that. And they said, "Well, can you do some analysis on that? " And they go, "Well, we already built it. " They went, "What?" So we built a machine learning model, an AI-based machine learning model to do inferencing, straight hardcore inferencing that you're probably all familiar with to analyze those pictures and say, "Oh, right there, look at that.

(29:06):
" And so we took something that took them weeks and we do it in seconds. So that's a very practical example of the capability. And there's many other examples of having that compute capability on board and what it can do for you. So NASA was so excited about that. We got the Team Flight Award, which we shared with a couple of our partners as well. And then again, you can see right here, the Spacebourne computer to flying, that's 2.5, in fact, flying in the Columbus module. All right, that's what we've been doing on the ISS. By the way, did I mention not only does Hewlett-Packard Enterprise have the most powerful computers in space, we also have the most powerful computers on planet Earth too with our supercomputing. Just want to make that point. We're kind of the intergalactic leaders of compute. All right. Okay. This leads us in and our work on the IS SAS led us into meeting some really interesting people and different organizations.

(30:05):
And our partners, we were introduced to a company called Astrolab. And so you can Google Astrolab, you can look them up on their website, astrolab.space, and they are building rovers and they are building rovers to explore the surface of the moon. Now there's a progression of missions here. This is called, we have FLIP, we have Flex, and then we have the lunar terrain vehicle. So it's a progression as we go. We are flying in the summer and late fall on the FLIP mission. So the FLIP mission, and you can Google this one as well, is called GRIFIN one. Griffin one. That is a NASA CLIPS mission. CLIPS is a program that NASA uses to encourage the exploration of the surface of the moon. So this is a NASA flight. It is going to fly on a ... We're going to take off, it's going to be a little bit like a traditional, what you might think of as from Apollo mission.

(31:00):
So it's a SpaceX mission, SpaceX Falcon Heavy goes up. Then a lander disengages from that rocket. Now that landing company is called Astrobotic. There's a lot of Astros, different company. There's a lot of Astros in this space game. So Astrobotic will separate from the Falcon heavy, fly to the surface of the moon, lander lands. This little buggy drives off the back of it. Now this little buggy will be exposed to the vacuum of space during that entire journey. So the completely robotic mission. And inside that little buggy is a spaceborne four compute blade. So this is a spaceborne four compute blade, low powered CPU, low powered GPU in here as well, and one terabyte of Kyoksia memory. This guy is actually bolted onto the radiator inside the flip, and this is conductive metal. So I'm holding the cooling system of this guy. So it radiates its own heat out.

(32:02):
And we actually have layers of conductive material in between the different cards that are inside of this. And it plugs in with an ethernet cable. Just a good old ethernet cable plugged right into the flip. Here, pass that around. Nobody steal that. Pass it around. Take a look at it. Feel the weight of that. Please don't steal it. That's the only one. Okay.

(32:25):
Flex is the size of a small truck. Now, flip is going on, Griffin. One, Flex is part of the relationship with SpaceX and SpaceX has three missions contracted as part of the Artemis mission. One is pure cargo. Two is going to be holding astronauts. So we're not sure which one we're going to be on. And we're going to have a larger system N array in the Flex Rover as well. And then that leads to the lunar terrain vehicle contract program, which we believe Astrolab is going to win a lion's share of. And that's a fleet of Rovers, a fleet of little trucks to build the infrastructure on the surface of the moon. Ladies and gentlemen, this is happening, and it's happening now, and it's happening in our lifetime, and we're all going to get to watch it and be part of it. Okay.

(33:12):
This is the Flex Rover. So AstraLab was competing. There was a big competition for this contract, about 30 different competitors. They were down selected to three competitors, and that down selection allowed them to build this prototype and this system. It's pretty cool. And notice that it's kind of like a cargo ship. It carries a mezzanine of cargo. So it literally drives over the cargo box, picks it up, carries it along, drops it off. This bad boy's going to be actually coming out and you will see this. You will see this. It will be set down from an elevator on a starship. SpaceX Starship, this flies on Starship. And you're going to watch that whole progression of it coming down. And you heard it here first. And it's not going to be science fiction. It's going to be reality. It's going to be our shared reality.

(34:07):
But why are we doing this just because it's absolutely cool? Don't you agree? Don't you think it's just pretty cool to do when you want to be part of this project? We're also doing it because it makes good business sense. Remember that edge example and the gloves and not shipping large datasets. That's what it's all about. PWC says that there's going to be a $170 billion lunar market. Now that's infrastructure, that's rocketry, that's all these different things. And this study is slightly old. My numbers are probably understated now, but it's also identified $12 billion in data acquisition. That's an addressable market. Hey, we're cool people at Hewlett-Packard Enterprise, but we're also business people. So we're going to get out there and we're going to take advantage of it. And you can see that there's all these different studies and different organizations really acknowledging that the space economy is just ... It's taking off, literally.

(35:02):
And look, and of course, all of these well-known individuals are now all piling on the notion of orbital data centers. Hey, guess what, ladies and gentlemen, we kind of have an orbital data center right now. So we're ahead of the game and we intend or we hope to continue to extend and work on this. Now, by the way, in orbit, we actually did the first large language model, AI-based large language model. I already talked to you about the inferencing engine, and then we have done a whole series of different experiments, and we're going to be doing the first AI experimentation and work on the surface of the moon as well. All right, so let's go take a rocket ship ride. This is the launch of Spacebourne Computer 2.5 in January 2024. One second.

Video Narrator (35:52):
Minus 10, nine, eight, seven, six, five, four, three, two, one, H engine. And with no stations, go falcon. Falcon nine and signus begin their flight, taking aim on the international space station.

Norm Follett, HPE (36:21):
So this is the Signes module which is now separated from the Falcon.

Scott Meltzer, Comedy Industries (36:24):
Robotic arm is two meters away from the spacecraft.

Norm Follett, HPE (36:27):
That's the Canadian arm that reaches out, pulls the spaceship in for docking. Space porn two, dot five, being installed. This is like the ultimate IT call, right? Service call. He's floating upside down. He's installing this in the ceiling of the Columbus module. It weighs about 125 pounds on earth and he's got two of them. Pretty cool. So you can find a URL, can find information there as well. Couple of tidbits in what you just saw there. So the Cigna spacecraft, and I mentioned early on that we flew on a SpaceX mission, pure SpaceX, and then we flew on a pure Northrop Grumman mission. Well, Northrop Grumman's rocket factory was in the Ukraine, so they couldn't find any more rockets. So what ended up happening is that launch that you just watched was a hybrid. It was a SpaceX moving a Cignas capsule. And they're competitive companies, but NASA literally said, "No, we got to get stuff up there so they find a way." So the space industry is full of partnerships, competition, but in the end, it's human beings trying to get something very, very special done.

(38:03):
And that's a pretty good example. And by the way, Hulu Packard Enterprise is the only company that's been on every variant of the rockets that go up to the space station. This is the Flip Rover, gives you a little bit of an idea of the size of it. This happened 10 days ago, and I was there for that. This is our lead mechanical engineer, Dave Peterson. He's actually bolting that system onto the chassis. We are literally bolted in right now for the mission. Now, we're going to go through some electrical testing and some overall certification, but it's real. It's happening. It's happening soon. And then this is the representation of the As Robotic Rover and then a representation of the FlipRover driving off. And ladies and gentlemen, thank you so much. Remember Hewlett-Packard Enterprise, the most powerful computers on the planet, the most powerful computers in space, and soon the most powerful computer to ever be in the moon.

(38:57):
Follow us, join us, send us emails, join the mission. Come on, let's go to the moon.

Scott Meltzer, Comedy Industries (39:02):
Hey, thank you very much, Norm. Thank you, Norm. Thank you, Igor. That was fabulous. We are going to do our drawing right now. Take your green tickets. If you have more than one, place them in numerical order so we can find the winning ticket fast right after we do our drawing. If you would like to learn anything more about the Spaceborne computer, you can see an actual mockup of ... It's a reproduction of the chassis is right over there in the corner. You can learn about our acceleration playbooks over here and you can see all the HPE demos right here in our booth.

Katrine Spang-Hanssen, Comedy Industries (39:31):
And if you would like to see the acceleration playbook right around the corner, that's just a two minute demo of our HPE networking, how it was powering up the writer's cup. Yeah,

Scott Meltzer, Comedy Industries (39:44):
That was impressive.

Katrine Spang-Hanssen, Comedy Industries (39:44):
All right, for right now.

Scott Meltzer, Comedy Industries (39:46):
The winner is. The

Katrine Spang-Hanssen, Comedy Industries (39:46):
Winner is. We're looking for the person holding ticket number 971272.

Scott Meltzer, Comedy Industries (39:54):
That's nine, seven, one, two, two, seven, two.

Katrine Spang-Hanssen, Comedy Industries (40:00):
Going once.

Scott Meltzer, Comedy Industries (40:02):
Nine, seven. Seven, one, two, seven, two, going twice. Twice is enough. Move on. If you can't find it in that much time, you have too many tickets and that's your fault.

Katrine Spang-Hanssen, Comedy Industries (40:15):
Here we go.

Scott Meltzer, Comedy Industries (40:16):
The real winner.

Katrine Spang-Hanssen, Comedy Industries (40:17):
The real winner. 971352. That's

Scott Meltzer, Comedy Industries (40:21):
Nine, seven, one, three, five, two, so close.

Katrine Spang-Hanssen, Comedy Industries (40:28):
Nine,

Scott Meltzer, Comedy Industries (40:29):
Seven,

Katrine Spang-Hanssen, Comedy Industries (40:29):
One, three, five, two. Here we

Scott Meltzer, Comedy Industries (40:31):
Go.

Katrine Spang-Hanssen, Comedy Industries (40:31):
Nine, seven, one, three, five, two.

Scott Meltzer, Comedy Industries (40:33):
How close were you? Oh, not even close. You just ... Oh, so close. You can get close to it. You can get close to it. There you go. You're close to it. Okay, here we go.

Katrine Spang-Hanssen, Comedy Industries (40:43):
All

Scott Meltzer, Comedy Industries (40:43):
Right.

Katrine Spang-Hanssen, Comedy Industries (40:44):
All right. The real winner. The

Scott Meltzer, Comedy Industries (40:46):
Real winner.

Katrine Spang-Hanssen, Comedy Industries (40:48):
971310.

Scott Meltzer, Comedy Industries (40:50):
Nine. Seven, one. Three, one, zero. Right over there. All right. We

Katrine Spang-Hanssen, Comedy Industries (40:56):
Have a person way back here. Let's

Scott Meltzer, Comedy Industries (40:58):
Check this right here.

Katrine Spang-Hanssen, Comedy Industries (40:59):
All right. And we have found the winner. Clap us if you were the winner and we'll all feel great about ourselves.

Scott Meltzer, Comedy Industries (41:04):
If you'd like to learn anything more, follow me right over here. You can see our acceleration playbooks demo right over here. You can learn all about HPE solutions for accelerating telecom networks and monetizing your telco networks right over here.