Ep25: MI&S Datacenter Podcast: Talking HPE, AI, HPE & Pure Storage, Zscaler, Virtual Animals, Oracle

Welcome to this week’s edition of “MI&S Datacenter Podcast” I’m Patrick Moorhead with Moor Insights & Strategy, and I am joined by co-hosts Matt, Will, and Paul. We analyze the week’s top datacenter and datacenter edge news. We talk compute, cloud, security, storage, networking, operations, data management, AI, and more!

2:20 HPE Discover 2024 Insights
11:59 AI For President
18:40 HPE and Pure Storage Close Out Conference Season
27:09 Zscaler Zenith Live 2024 Insights
33:10 Virtual Rats
39:15 OCI Is The MultiCloud

HPE Discover 2024 Insights

https://x.com/WillTownTech/status/180…

AI For President

https://www.nbcnews.com/tech/tech-new…

HPE and Pure Storage Close Out Conference Season

https://www.forbes.com/sites/moorinsi…

  / urn:li:activity:7208893987951136768  

  / urn:li:activity:7209593186044846081  

Zscaler Zenith Live 2024 Insights

https://x.com/WillTownTech/status/180…

Virtual Rats

https://doi.org/10.1038/s41586-024-07…

OCI Is The MultiCloud

https://moorinsightsstrategy.com/rese…

Disclaimer: This show is for information and entertainment purposes only. While we will discuss publicly traded companies on this show. The contents of this show should not be taken as investment advice.

Matt Kimball: Good morning, good afternoon, good evening, good night. Good, whatever. Welcome, welcome, welcome to the, wow, which one is this, guys? This has been a, we’ve had a lot to our 25th is it? 25th Data Center podcast.

Will Townsend: What anniversary is that, Matt? Is that diamond, sapphire, ruby?

Matt Kimball: I think it’s the reconsider your relationship anniversary. I think that’s what it’s called. But anyway, welcome. My name’s Matt Kimball and I am with my good friends and cohorts and co-hosts, Will Townsend and the estimable, estimable something like that, Paul Smith-Goodson.

Paul Smith-Goodson: Whatever.

Matt Kimball: Hey, we got a lot to talk about today, but before we get going, a couple of apologies and a disclaimer. The apology is, sorry we’re a little bit late getting going. I had to take my vitamins and those took a couple minutes and then I screwed up when I was starting the broadcast, so we had a little bit of a hiccup. We’re back. Thank you. And the disclaimer is this, if you use our advice for investing or any kind of financial purposes, see a therapist because you are crazy. Do not take anything we say as credible. No, I’m kidding about that. But don’t take anything we say as advice around financial investments or otherwise in companies. We’re just giving you our opinions and we’re having fun while we’re up here.

So just want to throw those out there. So jumping right in, we got a lot to talk about. HPE, HPE again, AI, I know who that is. Pure Storage, really good friends in the storage game. Zscaler, virtual, oh gosh, virtual animals, I don’t even want to know. And finally on the heels of that, Oracle, and if you’ve seen us more than once, you know who’s doing virtual animals and you probably know who’s doing AI. But anyway, and again, don’t take what we’re seeing as advice in any way, shape, or form. Let’s jump right in because we want to fly through this and let people get on with their 4th of July breaks. We’re going to talk about HPE Discover 2024 insights with the king of networking and security, Senior Will Townsend. Go get them, Will.

Will Townsend: Thank you, Mr. Kimball. So Discover, it was a spectacle. The company is the first to have conducted a keynote within the Sphere, and if you haven’t been to the Sphere, it is amazing. I’m looking forward to Black Hat in early August where I’m actually going to go see the dead experience with my wife, my brother and his wife, and it should be a lot of fun. But I did publish a research note with some of my insights. I know, Matt, I want to get your take on things as well. You and I were there with our esteemed leader, fearless leader, Patrick Morehead. And I tell you what, I came away pretty impressed. And in my research note I speak to two things. One, around what HPE is doing from a networking perspective. So traditionally Aruba was allowed to do its Atmosphere event separately. This year marked the first time where they integrated the Atmosphere event into the HPE Discover agenda and that’s for good reason from my perspective.

At the event, Antonio did speak to the importance of networking. And it’s something that I think gets lost in all of the generative AI discussions around one of your areas of coverage, which is obviously compute, GPUs and MPs and that sort of thing, but you’ve got to have robust connectivity to move all this stuff around. And what I continue to be impressed with is that that HPE, relative to say companies like Dell and Lenovo, they’re making a very, very focused investment in networking. They’re not just using Broadcom silicon and putting it out there. They’re actually doing a lot more. And what’s also impressive, this hasn’t closed yet, but certainly it’s no new news that HPE is acquiring Juniper Networks and with Juniper comes Mist and all the AI goodness there. And if you look at what Aruba has accomplished just with machine learning and AIOps over the years, the combination of both companies together is going to be pretty powerful and it’s going to fill a lot of gaps.

So when you look at routing, switching wired, and wireless, both companies coming together provide a complete solution. I’ll also add that Juniper brings some custom silicon capability as well. And so the combination of the two, really the only other company that can rival the depth of this combination would be Cisco in my opinion. So from my perspective, networking could really serve as a key differentiator for the company, especially as generative AI continues its momentum. This wasn’t announced at the event, but just a couple of weeks prior, HPE Aruba took the wraps off a new private 5G solution. They’re a little bit late to the category, but what I really like about what they’re doing, it’s an end-to-end solution. So Athonet was an acquisition that provides core infrastructure for cellular connectivity that is integrated with HPE branded radios. And when you look at also the software stack, this is all HPE.

So when you look at other private 5G solutions, a lot of times the company will provide the core infrastructure, but the radio tends to be a pickier flavor and those companies are typically JMA Wireless, Airspan for some time, but NEC and others. And again, what I really like about what they’re doing with private 5G is that it’s an end-to-end solution. They’re very focused on the usability. So they’ve integrated Wizards to make the deployment very simple and they claim that a customer can take their full stack and deploy a private 5G network in less than 30 minutes. And if that’s the case, I think it’s a total game changer. Those are my two big networking observations.

And I want to get your input, Matt, maybe we’ll start first with labs. I hadn’t had a lot of interaction with the Hewlett Packard Labs team, but you and I and Patrick, we had a couple of different conversations with executives and fellows and I was really applied with what they’re doing with applied research with respect to data center sustainability and also just broader generative AI applications. And the sustainability piece is huge because the compute infrastructure is consuming tons of power and so we’re going to have to get a lot more clever or the tech companies are going to have to get a lot more clever about how they reduce the power envelope there. But I think that’s a great segue to talk about compute and maybe some of your additional thoughts on the labs effort.

Matt Kimball: Yeah, I’ll hit on the labs and save the compute for a couple minutes from now, but I’ve always talked about, or you and I were just talking about, I’ve always talked about, because it’s all about me, you and I were just talking about prior to the podcast, about HPE labs and kind of the innovation engine that HPE is and sometimes they don’t get enough credit for it.

Will Townsend: I agree.

Matt Kimball: I often kind of think of them as kind of a science project for the business world or for the IT world. And I don’t mean that in a negative way. Sometimes that can be kind of a negative connotation or association, but I mean it in a very positive way. HPE does a lot of things both in their labs and then bring to market that other companies catch up to and maybe mainstream. And they’ve done it over and over again. They’ve done it in terms of compute, they’ve done it in terms of security, they’ve done it in terms of cloud enablement and they continue to do it. And sustainability is another one of those areas. They have over 300 patents in the area of liquid cooling alone. So there’s a lot that they do that.

I think sometimes companies kind of come and they fast follow and they take and they quickly mainstream, but we forget that a lot of those routes are built into HPE. Real quickly on the severe thing, I thought they did a, it’s easy when you’re in the Sphere to overwhelm everybody and have everybody go, “Wow, that’s fantastic.” And sometimes you get so focused on, “Wow, that was fantastic of an experience.” That you lose the meat of what’s being told in a keynote. And I thought the creative team at HPE and CEO, Antonio Neri did a great job of using this Sphere to articulate and emphasize the messages that he was trying to drive across and not overwhelm or overtake or overshadow the message of what he was saying. Really good use of the Sphere to kind of bring all of that kind of storytelling home. I’ll shut up on that and let Paul weigh in because I’ll go on forever if I don’t.

Paul Smith-Goodson: Yeah, I’ll just make a couple of comments. HPE has got a lot of carrier grade expertise and they can use their experience in orchestration and automation provisioning and with the radio they can repurpose all of those capabilities for the private 5G networks. So I think it’s good.

Will Townsend: Yeah, I’m glad you mentioned that, Paul, because I didn’t talk about it. So they’ve operated their CTG group for many years being very focused on big public mobile network operators. They recently decided to divest that and they’ve sold assets away to really double down and focus on enterprise. And I completely agree with you, a lot of the work they’ve done around, for example, open RAN acceleration, OSS and DSS, to your point, automation and provisioning, I got to believe, although I’m just reading between the tea leaves, that they’re leveraging that and leaning into that to deliver this full stack solution. And at the end of the day, I think it’s just going to make it a lot easier for customers to deploy private 5G networks. But let me finish that with a little factoid that I learned about the Sphere, Matt, guess what the monthly power bill is for the Sphere in Los Angeles?

Matt Kimball: Oh my gosh, at least a couple hundred bucks.

Will Townsend: At least. So if you can believe it, with all the LEDs inside and outside, over $500,000 a month for the power bill. And if you remember, it was super ice cold in there and I just thought maybe it was the day that we were there, but no, I’ve talked to a lot of people that have been there, they always keep that sucker cool. And it’s probably because of all of, and LED lights obviously run pretty cool, but it’s got to be because of all the power that the thing is consuming. So they need to get with HPE and figure out a more sustainable power bill to run that place.

Matt Kimball: $6 million a year. Wow, that brings me back to Steve Austin. All right, speaking of Steve Austin, let’s talk about, I don’t know why I associated Steve Austin with Paul Goodson-Smith, but that’s cool. Let’s talk about our next topic, which can only come from Paul Goodson-Smith, Smith-Goodson. I’m sorry. Man, sorry, Paul. Let’s talk about AI for president. Kind of timely given last night’s debate. You want to jump in on that one, Paul?

Paul Smith-Goodson: After last night, it may not be a bad idea, timely topic, but there’s actually somebody in the UK, his name is Steve Endicott and he’s running AI as a candidate for the UK Parliament. Endicott’s got an AI company, but he’ll be on the, it’s AI Steve and he’ll be on the ballot with all the other non-AI candidates. It’s pretty interesting, maybe a wave of the future to have AI on the ballot or running some small functions in cities or something right now. It’s also kind of points to the direction we’re going with AI for someone to actually come up and have enough courage to put an AI model on the ballot. But he’s got quite a website. You can either be a creator or you can be a validator on this website and they vote for policies that they like and then they confirm how much, what percentage of approval certain policies have.

It’s quite an operation. So the Endicott ran for a local election back in 2022 and his politics, he’s pretty much a centralist. I would expect him to be maybe a little bit more on the liberal side, but his motivation is to push the government to, he wants to cut carbon emissions and he’s not really promoting his business, I think it’s very sad, but he’s likely he’ll be the first AI legislature in the world, I guess. There’ve been some other candidates using AI for decision-making. One of the Danish parties used some type of AI to derive a platform for helping with decisions. So we could end up with an AI boss or something and maybe Pat Morehead is going to be replaced eventually by AI or we could be too. It’d be interesting

Matt Kimball: If you’re listening, Pat, that was Paul Smith-Goodson who said that.

Will Townsend: Not Will Townsend or Matt Kimball.

Matt Kimball: Is there another kind of point he’s making here about making or building AI more into the legislative or policymaking kind of process?

Paul Smith-Goodson: He wants it to really be democratic and you can log on to AI Steve and express your opinions or you can offer a policy or something that should be enacted and then everybody else can vote on it. And it’s basically if you get 50% approval on some type of policy, then they push it. It’s just a way of expanding democracy.

Matt Kimball: Is that ripe for abuse from bots and other kind of terms of-

Paul Smith-Goodson: I didn’t read anything about that, but I would assume that there should be some way that there you might be able to [inaudible 00:15:45] people to comment.

Matt Kimball: It’s interesting. It’s kind of, not scary is the wrong word, but I guess it’s kind of where we’re headed.

Paul Smith-Goodson: It’s really interesting. I mean, one of the areas I’m watching is AI and robotics and this is certainly an area that could feed into that in terms of experience.

Matt Kimball: Yikes. Will, any final thoughts before we move on to HP part deux?

Will Townsend: This is interesting. I wasn’t aware of it and if you sort of think about the voice of the people, I mean Paul, to your point, I mean it could really facilitate that. And so much of politics is just about, I mean, Dennis Miller had a funny riff on the POTUS and it’s like, why on earth would anyone spend billions of dollars to win a job that pays 200,000 a year? Now I do believe the POTUS has paid 500,000 now, so that dates my age. But yeah, so it’s often politicians are not necessarily focused on their constituency and maybe AI could level that. So I think it’s interesting,

Paul Smith-Goodson: You can also go into AI Steve and ask questions about policies that he’s pushing or about other areas.

Matt Kimball: Yeah, I think some of the guardrails you have to put up, it’d be interesting in the US, do bring this back home to the US, if you each congressperson had kind of an AI, I don’t know, constituency assistant, and if you could put the guardrails up that assured that it was just your constituency that was able to kind of interact with you. So Paul can’t interact with a congressperson from Massachusetts or something like that. That and if you could assure fuller participation from your base, the other concern is you’re hearing from people who are tech-savvy or people who, not necessarily the voice of everybody, it might be the voices of the few, but the more vocal that are kind of being represented.

Paul Smith-Goodson: Endicott sees this as kind of a reinvention of democracy.

Matt Kimball: Yeah, yeah. But the question is, are all the participants in that democratic process-

Will Townsend: Represented.

Matt Kimball: Represented or not.

Will Townsend: Talk about AI PORUS potential.

Matt Kimball: I think we’re going to get into a political fight here.

Paul Smith-Goodson: Off the normal path of AI, kind of an interesting use of AI.

Will Townsend: For sure.

Matt Kimball: Yeah, so speaking of AI, because no one ever talks about it, let’s talk about Discover just a little bit more in Pure Storage Accelerate. I’m going to go quickly because I’m hitting on two topics in one. Will, you did a great job of hitting on what HPE talked about at Discover from a networking and introduced compute as well. What I found, and you hit on it was they did a, I thought Neri did a really great job of connecting compute storage, software acceleration and the movement of data, how it gets created, how it gets moved, how it gets stored, how it gets protected throughout the AI journey. He really did, I thought a fantastic job. I was a little bit surprised by how strongly he embraced NVIDIA, but everybody is. Dell did at Tech World. Lenovo will do it at their upcoming event in October.

Supermicro has done it even though they don’t have a customer event that I know of, they’ve done it. Everybody’s out there kind of embracing NVIDIA, embracing NVIDIA, embracing NVIDIA. So not that surprised in that he did it, but when I went back and I looked at reference configurations, I noticed that a lot of companies, for instance, are doing Gaudi-based reference architectures as well. Supermicro, Dell. HPE doesn’t have one. So to me it’s where the talking goes into execution that they really seem to have embraced NVIDIA’s following. Couple of quick takeaways, really cool platform, the DL384, which is a ARM based, it’s Grace, NVIDIA Grace. It’s a Blackwell MGX-based platform was announced, which I thought was really, really cool as well as a Blackwell and Xeon, the 380A platform announced. They also announced a virtualization stack, which I wrote a big write-up on this on Forbes.

I would encourage folks to go read it where they’re going. There’s a lot of dynamic stuff going on in the virtualization space. I wonder if they are doing this, HPE is doing this at the cost of, not the cost, but maybe bruising the egos or the relationships they have with folks like Nutanix, SoftIron, Red Hat, but I get what they’re doing. I think it’s great to have a bespoke cloud stack. And something that didn’t get talked about a lot that I think they should spend more time on is what they’re doing with OpsRamp, an acquisition they made, I think it was a couple of years ago, and how that is providing organizational-wide, infrastructure-wide observability and management. This is a tool I’ve seen and everyone has, Dell has iDRAC, Lenovo has XClarity, HPE has had OneView forever. OpsRamp is the one tool I’ve seen that is kind of an equal, how can I say it? It treats all infrastructure equally from a observability and management perspective. This really could be that single pane of glass. I’ll be interested to see what they do with that.

On the Pure front, I’ll ask for your input in a second, guys. On the Pure front, AI made simple. I love what Pure is doing. The days of having IT organizations with storage administrators, network administrators, server administrators, app dev folks kind of in silos is going away. It’s all one big bucket now. So there’s a big focus on making all of these components more simple. That’s what Pure has done for storage. I love what they’re doing with Evergreen. It’s a storage as a service offering and Evergreen/Forever, which is essentially kind of work with us and you’ll never have to worry about your storage strategy, file, block, object storage, all managed from one interface very easily can be stood up by developers as well as storage admins. Love what they’re doing. I think storage is often overlooked in the AI equation, which is funny because data is the most critical aspect or critical part of any AI strategy. Pure really is on the, just as HPE is on the leading front of innovation and as it relates to compute, I think Pure has done this on the storage front and I think there’s a lot of upside to what they’re doing. I’ll be curious to see how that messaging and product lands with its customer base, but fantastic and compelling offering. All right, I’m going to send it to the round table to give some feedback on what I just spit out in 20 minutes worth or seemingly 20 minutes of overview.

Will Townsend: I’ll jump in, Matt. So just to comment on OpsRAMP, I heard a lot about it at this Discover and that was new for me, because I don’t think it was as highlighted at prior events and maybe that was based on the need for further integration and that sort of thing, but it seeming, and I agree with you, if it delivers on all of its promises, it could be very, very compelling. And HP seems very bullish obviously about it given the amount of time that was spent discussing it and how it’s a part of their overall tech stack strategy. And yeah, I got to agree with you on NVIDIA. I mean, it sounds like it’s sort of a double down, but from my perspective, I really view what they’re doing as a partnership of equals, and I sort of spoke about this, about Cisco at Cisco Live when they spoke to their collaboration with NVIDIA, so relative to maybe some other companies out there.

And I think the private AI cloud, I think that’s a very, very compelling solution that they put together. I like what they’re doing, I’m going to lean into your swim lane a little bit, I love what they’re doing with computing clusters, the small, medium, large, extra large, it made me think about my t-shirt collection. And the fact that given CUDA and the maturity of CUDA and just what they’ve done with building their AI tech stack and allowing customers to sort of enter at a level that they feel comfortable with and sort of grow pay as they grow, I think that’s super compelling and it really matches really well with what Antonio spoke to you about just their server cluster strategy. So sorry if I leaned into your swim lane a little bit.

Matt Kimball: Don’t never do it again. No, you’re absolutely right. You’re absolutely right. I couldn’t agree more on the love the story around kind of t-shirt sizes and making it simple and consumable. That’s really what enterprise AI is all about. But I don’t want to steal from Paul, because I know he always has smarter thoughts than either of us have on-

Will Townsend: For sure.

Matt Kimball: You guys are the brains in this part of it. I understand that HPE’s going to integrate the embedded stack through GreenLake, is that right?
That’s right, yep. They’re standing it all up through GreenLake consumable and they wrap services around it as well, which I think is a smart move.

Paul Smith-Goodson: Yeah. So they can deliver it like a turnkey solution for enterprise and basically a cloud service. So it worked out pretty good.

Matt Kimball: And they’re really thoughtful about it, Paul, because what they did was they stood these things up as services and they said, “Wait a minute, when you think about this from a go-to market perspective, there’s a whole ecosystem of channel partners and GSIs, global systems integrators that are building practices around this. Let’s go to the WePros and the Accentures and so on and so forth, and leverage their expertise and their practices in delivering these out to the market.” So really I think it’s a smart move to kind of lean on that channel. It emboldens the channel and it’s a profit, good revenue opportunity for them. And it also makes it easier and more consumable for customers. But we could talk forever and we probably will if I don’t move. Let’s go to Zscaler, my friend. Again, Will you want to hit on what you saw at Zenith Live and kind of what’s going on?

Will Townsend: Yeah, so because I cover networking and security, I’ve got to do some due diligence with my security swim lane here. But I’ll tell you what, the last five weeks have been pretty crazy for me. Four visits to Sin City and Discover was the culmination of that. But prior to Discover, I did spend time with Zscaler, got to spend time with Jay Chaudhry, who’s the CEO, and a couple of things that I noted there. Number one, let’s talk about NVIDIA a little bit more. Why not? So Zscaler has been working with NVIDIA for a number of years and what they’re doing now is they’re sort of supercharging, Zscaler’s supercharging what they do from a digital experience perspective with two new AI, generative AI copilots that are powered with NVIDIA technology. And the first is the Zscaler ZDX Copilot with NVIDIA Morpheus, and this is a GPU accelerated AI framework.

And then there’s a second around NVIDIA NIM, and I’m sure you and Paul are much deeper on the NVIDIA tech stack than I am, but my understanding is with NIM, it’s leveraging a set of microservices to accelerate AI inferencing and training and that sort of thing. So it’s just making what Zscaler does from a digital experience monitoring perspective smarter. And then the other big announcement there was a collaboration with Google Chrome on infusing Zscaler technology and then also Zscaler ingesting Google Chrome technology to improve browser isolation. And there’s been this whole sort of, from my perspective, momentum that’s beginning to bubble up around enterprise browsers as a potential VPN eliminator. And so I think it’s quite impressive on the service. I won’t go into a lot of details for sake of time here, but I did post a Forbes article on this and I was really surprised. I got up to 40,000 views on this.

So apparently there’s a lot of attention on Zscaler. I mean, financially they continue to do very well. Their latest earnings were super rock solid. And what I really like about what they’re doing is you could almost argue that they are sort of leading the whole notion of zero trust. They were very early out of the gate with that. They operate the largest security cloud on the planet. And I really look at these partnerships with Google and NVIDIA as ways that they’re really sort of fine-tuning and refining what they do and just taking it, putting the Ferrari into eighth gear here. And yeah, it’s quite impressive to kind of track their success over the years.

Matt Kimball: Hey, real quick, Will, before we move on, do you see what Zscaler is doing as, is this kind of iteratively kind of different I guess than what you see in the market? Or is this a, really kind going big and super differentiated than other players?

Will Townsend: I think it’s more iterative than anything. I mean, obviously they want to get credit for gen AI and they began that journey at Zenith Live last year with a couple of, I guess early preview things. But what they’ve done is they’ve refined that. They’ve probably learned from early preview and now they’re GA with a lot of this. So I mean from my perspective, at least on the gen AI side, they’re getting credit where they need to, because if you don’t have a generative AI strategy, people are going to kind of go, “What’s going on there?” But that’s a long way to answer your question, but I see it as more iterative, but they continue to fine tune things and really position themselves very well. So, yeah.

Matt Kimball: Just one last thought on that kind of enterprise browser, you caught me on that one, caught my attention. It’s funny, it’s going to be hard to kind of parse enterprise browser versus consumer browser. I would imagine that that enterprise browser becomes a consumer browser over time, or even the protections you want within the enterprise and maybe even the guardrails you want within the enterprise are things that as a parent could see, well, not anymore, because my kids are too old, but still want to, you’d want to apply if you’re browsing from home or on your personal laptop as well. That’s interesting. I like that.

Will Townsend: Yeah, I mean, I think there are certain requirements within the enterprise that sort of supersede what the standard free Chrome browser provides. And certainly Google has done a masterful job with their freemium model getting you hooked on the freebies and then like, “Oh, I’m out of Google storage space.” And I started paying for the first time last year for additional storage there. And it would be interesting to see how they do that. I mean, certainly they’re what they’re doing with enterprise browsers. But I mean, you’re right. I mean, especially as for example, AI PCs continue to gain momentum and we start seeing models move at the edge of the network and there’s going to be a need for enhanced security and all of that. So I sort of agree with you that it may move that direction over time.

Matt Kimball: Interesting.

Will Townsend: Yeah.

Matt Kimball: All right, let’s go on to, I can’t even say this without chuckling a little bit, but Paul’s favorite topic, virtual rats. Paul, what on earth are you talking about virtual rats?

Will Townsend: Always going for the headline.

Paul Smith-Goodson: I mean, everybody needs a pet. It was actually a study done recently by Google, DeepMind, and Harvard. So it’s a legitimate study in a legitimate subject. So you guys shouldn’t make fun of rats. They used virtual rodents, I’ll use rodents, that may be less offensive for you guys, to provide insight on neural control of movement of living things. They’re trying to better understand how brain controls work and how it relates to a biological model of it and how it relates to the living thing. And this study gives an understanding of how that happens, how the brain controls a lot of the complex and movements relate to the neural activity within the animal. So what they did was they created a system they called motor imitation and control. And I’ve got a private theory on that. I think probably a bunch of scientists got together and probably spent 10 hours trying to figure out how they could come up with a name of the study that spelled out Mickey Mouse. But this one’s just called Mick.

They recorded this and they had a bunch of rats and they had kind of like a little circus ring around them and they had six cameras and they collected 3D poses of these real rats and the rats will run around the circle and they’ll stand up on the wall and try and get out. And so they captured a bunch of poses of the rats and they used that and made a virtual rat model in a simulator. And then they trained the artificial neural networks to control the virtual rat to try and copy the behavior of the real rats. So they measured, they had measurements they took of the rats in basically two areas of brain. One is called dorsolateral striatum and the other is the motor cortex controls. The DSL basically controls, it’s got a sinking control and a movement control and it’s pretty complex to measure, but somehow they put all this stuff together Anyway, they analyzed the 3D movements of the rats from the videos and they correlated specific behaviors with that virtual rat.

And the results were that the brain activity of these two areas that they were measuring was actually better predicted by the AI activity than actually anything that the rat was doing. So that kind of suggests that the brain areas are involved in actually calculating the forces that the rat needed to move and directions it needed to move in and stuff. And they found out that the artificial neural network activity actually matched the patterns across the brain of the rat. It kind of suggests, the conclusion of the study was that using these virtual rats can kind of help us understand how the brain relates to movement and control.

Now of course, you can extrapolate that into human control and things like that, but it could be used for things like paraplegics and good things too, as well as could be evil things involved too. It was a successful study. They had some theories and stuff they wanted to validate, and the study provided them with that. And it was predictive of how the neural areas acted with behaviors. So it was a good study. I think we’ll see more studies like this maybe on larger animals rather than rats. And I think it may translate into medical applications in the future.

Matt Kimball: Well, what about behavioral applications? I mean, could this also, especially with humans, I mean, I can see and it’s kind of scary when you think about it, it’s fascinating and there’s so much potential, but as somebody who wears a tin foil hat on the weekends, I think about how the minority report starts to become a real thing. Yeah, wow.

Will Townsend: Hell of a rat race.

Paul Smith-Goodson: You put the rats together with AI Steve, and we have something.

Will Townsend: Oh yeah, let’s get those two together stat.

Matt Kimball: I love that. All right, we’re jumping on guys. We’re closing this out, I don’t know about strong, we’re closing this out with OCI though, which is strong. The topic is strong, but I’m not. So OCI is the multi cloud. I’m going to go quickly for the sake of time. So Oracle, interesting, Oracle about two weeks ago announced their earnings and they undershot on their earnings. They did well, but they came in under expectations. Despite that, their stock went up like 15% overnight. And the reason why it went up overnight is because accompanying the less than great expected results were a couple of things. First, they announced OCI or Oracle database at Google Cloud, OCI interconnect, high speed interconnect between Oracle Cloud and Google Cloud and partnership with OpenAI to provide OpenAI services in OCI. I say all of this, oh, and by the way, right after that, they announced Oracle Autonomous Database availability at Microsoft Azure. I say all of this to demonstrate the strength of open cloud or open multi cloud capabilities and what that can mean for a company.

So what these things are, start with database at whether it’s Google or Azure, Oracle is actually taking its hardware infrastructure, its exascale database machine, and it’s OCI infrastructure and it’s putting it in the data centers of Google and Azure and it’s standing up OCI instances, database at Azure, a bunch of services at Azure and Oracle so that these OCI services run native within those respective clouds. So a customer who has all of their applications running in Azure or their applications running in Oracle or Google Cloud can easily, from their Azure interface console or Google console select to run these on the Oracle database infrastructure natively. They get built through Azure or Oracle or Google. They manage it through their Azure console. Everything is done through Azure, but it’s running on Oracle cloud infrastructure. So performance can’t be beat. It’s sitting within the data center, it’s right next to the data, it’s right next to the, or the data’s right next to the application. Security can’t be beat because it’s not going over any kind of interconnect. Manageability can’t be easier because it’s all managed from a single interface. Billing can’t be, and operational can’t be any easier because it’s all being built through a single mechanism. That to me is what multi-cloud is all about. It’s not saying, “Hey, you can run things here in there, but you’re going to have really big data ingress and egress fees.” Or, “You got to manage this through multiple interfaces.” It’s all being done and built simply.

That’s part one. Part two is this high-speed interconnect. So if for some reason you can’t implement that database at kind of implementation or there are other Oracle services you’re using, or someone like Azure or a company like Azure wants to leverage OCI to help improve its OpenAI services, they have these high 50 gigabit per second interconnects built between the data centers, super secure, kind of private tunnels if you will, that enable data and apps to be moved back and forth between the clouds from OCI to Azure or Azure to OCI, same thing for Google. No data ingress fee, no data egress fee, high speed, absolutely secure. So this again is true multi-cloud. You have things running in different clouds, but you’re able to get that native-like performance without sacrifice and without additional cost. I don’t understand why more clouds aren’t doing this. AWS lives in its ivory tower and I get they want to protect their territory, but to me this is the future of cloud computing. This is where everybody is going to go to. I’m going to leave it at that and let you guys weigh in.

Will Townsend: Matt, I totally agree. I mean, one of the big challenges with multi-cloud, and let me touch on security for a moment. One of the big challenges is typically when you look at GCP, when you look at Azure, when you look at the big public cloud providers, AWS, OCI, the underlying network architecture are different. And so when you’re having to not only you spoke to eliminating the ingress fees, which is huge, but from a security standpoint, that creates a lot of challenges for multi-cloud deployment. And honestly, when you look at enterprises, most are multi-cloud because they’re leveraging best of breed, they’re leveraging cost containment, that sort of thing. And when you do it the traditional way, it creates security vulnerabilities. And so then companies like Infoblox step in and provide DNS based services to sort of go address that. And hey, listen, Infoblox is a great solution. Scott Harrell, my buddy, former Cisco executive is the CEO there now. Great solution, but I agree with you. It’s like, “Wow, this is super, super efficient. And I’m glad you’re sharing this on the podcast because I had no idea.

Matt Kimball: Yeah.

Will Townsend: Good stuff.

Matt Kimball: It’s pretty cool. It’s pretty cool.

Paul Smith-Goodson: And other companies might be doing the same thing, so it’s going to basically reshape what the multi-cloud landscape look like over time.

Matt Kimball: Yeah, I think, yeah, Paul, the way you’re saying, I think competitively this is going to push, and they have willing partners in Azure and Google. It’s going to push AWS to open up a little bit and I think it’s going to push all of these cloud providers to start working more efficiently and seamlessly with each other and find differentiation in other ways, but we’ll see. I talked to Leo Leung at OCI, he’s done a great job and Karan at OCI as well, he runs the shop. They’ve done a great job of doing this and kind of building these partnerships. It’s going to be fun to see where it goes. All right, so listen, we wrapped up on this. In the spirit of the 4th of July holiday coming up, I have a quick question, Paul, eyes up front, I have a quick question for you guys.

Will Townsend: Eyes right here, *censored*.

Matt Kimball: Quick round of trivial kind of question for you both. So given the 4th of July right around the corner, give me your top two or three 4th of July or patriotic themed movies that you’ve seen. This could be from the ’70s, ’60s, ’80s, new generation. Who cares? Top three. Will, you go first.

Will Townsend: I go first, okay. I got to say Born on the Fourth of July.

Matt Kimball: Okay.

Will Townsend: Tom Cruise. Man, I’m struggling. It’s 4th of July movies. I want to think of a rom-com. Isn’t there a Sandra Bullock rom-com that’s probably with Matthew McConaughey and 4th of July, I don’t know. And isn’t there a scene in Forrest Gump, which is one of my favorite movies, and by the way, one of my favorite movies I like to watch with my oldest daughter, isn’t there a scene there during the 4th of July, I think? So I’m going to go with the non-descriptive rom-com, Sandy Bullock, Born on the Fourth of July and Forrest Gump.

Matt Kimball: Yeah. And just so I should have made, it could be patriotic. It doesn’t have to be tied directly to 4th of July. So Forrest Gump certainly plays into that kind of. Yeah, very good.

Will Townsend: Yeah.

Matt Kimball: Mr. Smith-Goodson, what do you got for it?

Paul Smith-Goodson: Avatar, demonstration of military power. And the Tom Cruise movie the jets. I forgot the name of it.

Matt Kimball: Oh, damn it. Okay. Yeah, good.

Will Townsend: Top Gun.

Paul Smith-Goodson: And Alien and Avatar.

Matt Kimball: Alien, okay.

Paul Smith-Goodson: Yeah.

Matt Kimball: That’s interesting. Okay.

Paul Smith-Goodson: Yeah, in space.

Matt Kimball: Good choice.

Will Townsend: Mickey Mouse.

Paul Smith-Goodson: You guys-

Will Townsend: All right, Matt, how about you, man?

Matt Kimball: You stole two of mine. I actually had four, two of mine. One was Top Gun, so I’ll go with Top Gun Maverick, especially since I saw it on the 4th of July recently. I had Forrest Gump, but that’s okay. I’m also going to go with The Patriot, Mel Gibson.

Will Townsend: Yes. Great movie.

Matt Kimball: Yeah, fantastic movie.

Will Townsend: Great movie.

Matt Kimball: Yeah. So yeah, those were mine. And I would encourage everybody to sit down and watch one of those movies. They’re all very entertaining. I’m not sure where Alien came from, but that’s all right.

Will Townsend: Hey, I’m going to throw out two more, Matt, because we’re on the subject, Air Force One. I love the scene when Harrison Ford’s like, “Get off my plane.” I mean, that is freaking awesome. And who doesn’t love Independence Day with Will Smith before the Chris Rock slap. Who doesn’t love Independence Day? I’m going to go watch Independence Day right after the podcast.

Matt Kimball: Pretty CB. That was a great one too. Good call. I love that. All right, hey listen, just before we close out, I have to say it one more time. We do this for entertainment purposes. We do not do this to guide investment policies or decisions, so please don’t take anything we say as kind of professional advice. We’re just giving you our opinions. Thank you for joining us. This was a great conversation and sorry for the hiccup at the beginning and I guess we’ll see you in a couple of weeks.

Will Townsend: Yeah. And happy 4th of July to all of our viewers.

Paul Smith-Goodson: Let me just add the caveat, I didn’t mean to give the DNC any ideas about using AI Steve.

Matt Kimball: Finally, guys, if you like us, make sure you click the like button. If you don’t like us, too bad. But thank you for joining us and we’ll see you in a couple of weeks. Guys, have a wonderful week, 4th of July and be safe out there. I want to see all 10 fingers next podcast to make sure you guys didn’t do anything stupid. See you guys.

Will Townsend: Bye.

Patrick Moorhead
Founder, CEO and Chief Analyst | + posts

Patrick Moorhead is the founder, CEO, and chief analyst of Moor Insights & Strategy. His big-picture view of technology is grounded in more than 20 years as an executive leading strategy, product management, product marketing, and corporate marketing functions at NCR, AT&T, Compaq, and AMD. He has shared his expertise in areas from silicon to infrastructure to enterprise SaaS and everything in-between in thousands of national broadcast appearances (CNBC, Yahoo Finance), articles (Forbes, CIO), research-based analyses, and podcast episodes. Today, he has 100+ CXO-level advisory clients and is often ranked the #1 technology industry analyst by ARInsights.

Will Townsend
VP & Principal Analyst | + posts

Will Townsend manages the networking and security practices for Moor Insights & Strategy focused on carrier infrastructure providers, carrier services, enterprise networking and security. He brings over 30 years of technology industry experience in a variety of product, marketing, channel, business development and sales roles to his advisory position.

Matt Kimball
VP & Principal Analyst | + posts

Matt Kimball is a Moor Insights & Strategy senior datacenter analyst covering servers and storage. Matt’s 25 plus years of real-world experience in high tech spans from hardware to software as a product manager, product marketer, engineer and enterprise IT practitioner.  This experience has led to a firm conviction that the success of an offering lies, of course, in a profitable, unique and targeted offering, but most importantly in the ability to position and communicate it effectively to the target audience.

Paul Smith-Goodson
VP & Principal Analyst | + posts

Paul Smith-Goodson is the Moor Insights & Strategy Vice President and Principal Analyst for quantum computing and artificial intelligence.  His early interest in quantum began while working on a joint AT&T and Bell Labs project and, during 360 overviews of Murray Hill advanced projects, Peter Shor provided an overview of his ground-breaking research in quantum error correction. 

Recent Posts