Grey Beards on Systems - 165: GreyBeard talks VMware Explore’24 wrap-up with Gina Rosenthal, Founder&CEO Digital Sunshine Solutions

Episode Date: September 9, 2024

Gina and I were at VMware Explore and we discuss some of the changes we saw there and what's the latest functionality coming out that matters to the enterprise.

Hey everybody, Ray Lucchese here. Welcome to another sponsored episode of the Greybeards on Storage podcast, a show where we get Greybeards bloggers together with storage assistant vendors to discuss upcoming products, technologies, and trends affecting the data center today. We have with us here today longtime friend and fellow podcaster Gina Rowentz-Senthal, founder and CEO of Digital Sunshine Solutions. We were both at VMware Explore Conference last week in Vegas. So, Gina, as this is your first time on the show, why don't you tell us a little bit about yourself and what you thought interesting at last week's VMware Explorer conference?

Okay. Hi, Ray. Thank you so much for having me on. I am honored to finally be on the Greybeards. That's exciting. About myself, I've been in the storage and virtualization infrastructure industry for quite a while. The last company job I had was at VMware, and now I run an agency that does fractional product marketing

for people in the same industry. I also have a podcast called Tech Aunties, and you'll have to be on Tech Aunties with us one day soon. I'd love to. That's cool. And you'll have to be on TechAunties with us one day soon. I'd love to. Cool. So, yeah, I was really happy to go to VMware Explorer. I went because my friend Tony Foster and I were going to redo our session that we did before the pandemic that had a horrible name. It was because nobody was really saying AI yet at that point. So we

called it HPC, machine learning, deep learning. I don't know what all else we called it, but it was basically about how an AI workload works and how you can virtualize it. And so we redid that this year and talked about a little bit how you can virtualize it on VCF. And I know there was a lot of sessions on that. I didn't get to go to a ton of sessions. I think I only made it to one because I was working in the hub in the community theater, helping them record all the sessions there. And that was kind of great. The sessions were, the community sessions are always so good. People talk about such relevant,

cool, interesting things. And right next door to us, they had the VMware code folks. They were packed just like our sessions were mostly packed all the time. Yeah, I mean, I just kind of walked through the hub and it seemed like it was pretty busy every time I walked through there.

I was sort of surprised with. I think there were just really good talks. That's what I noticed. You had the normal VMware communities, people talking things from VMware. But there was a lot of AI. There was a lot of security, a bunch of people giving security presentations. They were just really great, in-depth, solid presentations. I really wish I could have gone to more sessions, but I didn't feel like I missed a done.

I went to a couple of sessions on Monday on Edge and Gemfire and a couple of other things, vSAN, of course. I thought they were pretty well, you know, they weren't actually packed, but they were certainly full of people and stuff like that. Yeah, the one I went to was completely full. It was on cloud economics. Oh, interesting. So it was a good presentation. Yeah, I'm sure yeah i think that's the thing about vmware we can talk about this later maybe a little bit i know people are

trying to figure out with all the changes coming up how vmware or can they make vmware still work in their infrastructure but vmware, it's always amazing to me how solid the innovations are and how it's keeping pace with all of the hardware innovations and chip innovations. And their features just kind of line right up to them. I don't know. Did you see the same?

Maybe you don't agree with that and you should tell me more. No, I agree to a large extent. I mean, there was a lot this time for the analysts, at least on private AI, you know, foundation, which is NVIDIA partner solution that they have. And, you know, in the past, they talked about the private AI as a reference architecture and the private AI foundation. This time is more on the foundation aspect of it with NVIDIA. And there's lots of stuff to talk about. But, you know, the hardware-wise, they started talking about memory tiering, and they mentioned GPU HA, which I thought was kind of interesting.

And I'm happy to talk about that if you wish. And they were touting DRS as being the reason why they're a better AI environment, which is kind of interesting because DRS has been around for decades. Ever. So I'm a little confused about what was the use case they were using for? They mentioned that the challenge with orchestrating machine learning ops and things of that nature is that occasionally resources are not available or servers are down. And how do you manage that sort of variable environment or fluctuating environment? And DRS, to some extent, was meant for that, right?

I mean, it was built for that. And what they were kind of implying was that, OK, yeah, there are these orchestration solutions out there, but they don't have our 10 years of experience with DRS and things like that. And Chris mentioned GPUHA. So what does GPUHA mean? And to a large extent, it's sort of like you've got these GPU users. You've got the mainline inferencing. You've got training.

You've got the data scientists. They're all consuming GPU resources. But they certainly kind of have a prioritization scheme. So the inferencing, because it's enterprise, you know, it's the highest priority. The training is probably the next highest priority. And the lowest priority are these data scientists out there playing around with models and things of that nature. So with GPUHA, if a GPU goes down on inferencing, they'll take it away from the data scientists. Okay. All right. the data scientist. Okay.

All right. That makes sense. Yeah. Well, not to the data scientist, I'm sure, but. Yeah, yeah. Exactly. But I mean, if you want, you know, the enterprise is dependent on inferencing, not necessarily the data scientist toy, you know, solution that he's working on at the moment. Yeah.

And it's the inferencing is the one that takes longer too. Yeah, well, I mean the training takes a long time and stuff like that. Yeah, but you don't want it interrupted. Yeah, you don't want, you know, you've got customers on the line using your AI and whatever solution you've buried it into and they're doing all this inferencing

so, yeah, you want to keep those guys happy and well fed and all that stuff. Yeah, that makes sense. That's kind of what our talk was just about that. It was way more basic. But the talk was that AI is just a workload. It's not something magical.

And you have to be able to manage those workloads and you have to put them on the best available architecture. So part of that includes managing if there's a failure. So that makes total sense. Yeah. Yeah. Now, like I said, it's this private AI solution. So we should probably talk about VCF9. First question is why the number nine? I mean, it was from the analyst perspective, we were all kind of wondering what the story is here, but it turns out, you know, the challenges with VCF in the past, 5.1, 5.2,

5.3, was that it was not exactly up to snuff with the rest of the functionality, vSphere, vSAN, that sorts of things. So they went to a different way of working, I guess, because now everything is kind of tied to the VCF release. So when VCF 9 comes out, all functionality will be at the current level under VCF. But vSphere is not at 9 yet. It becomes a platform, you know, I would say the platform god, but that's not the right term. Platform container. No, it's like the locomotive behind which all the other functionality runs with, you know, and stuff like that. And they packaged SGDC.

They packaged a lot of functionality in BCF, which had been not necessarily all part of vcf in the past so does that mean they're going to deprecate the other software type like will they deprecate v sphere and now it's vcf no they've got a v sphere uh foundation solution right and they got two right they got vcf9 and v sphere uh vmware foundation or something like that, VVF and VCF, that they support and license and stuff like that. And it's just, vSphere doesn't have a lot of the functionality, automation capabilities that VCF comes with. It makes sense.

I mean, it makes sense to have everything from a product marketing standpoint specifically. If that's what you're marketing versus all the individual pizza parts and then having to keep up with all the different SKUs and who gets what, that makes a lot of sense. It makes a lot of sense just from, I remember internally, nothing was tied together. You had all these different business units and they did what they wanted to do, whether it made sense for the overall company or not. If it made sense for their particular part of their particular product. It made those features sort of very nimble and very, you know, future focusing and trying to get those things out as quickly as possible. But in the end, I mean, the customer wants the package. They don't want these point solutions here and there, I think. And that's where by, you know, integrating themselves behind VCF as the major local mode of release.

And then it makes it just so everybody's talking the same game. Everybody's trying to work better together, which is always a challenge in these sorts of environments. Yes. It's just that I see as being better for the customer for sure. But, you know, let's talk about storage for a minute. Because like, what if I don't want vSAN? So now if I pay one price for VCF, I can use anything in that bucket now.

It's more than a bucket, but you know what I mean. It's more like a cloud or something like that. Yeah. It's huge. And you're right. So these customers are now buying the whole package, and they get all these solutions, vSAN being one of them, NSX, obviously, et cetera, et cetera. And they can take advantage of it or not, depending on what their requirements are.

I was talking to a couple of customers Monday morning at breakfast, and, you know, they were all looking at vSAN, but they had all had, you know, they're kind of in the VCF bucket. Well, two of them were VCF and one was VVF, but they already had storage, right? You know, primary storage from vendors, et cetera, et cetera. And they were thinking about implementing vSAN for a couple of clusters, I guess, to try to work with it and see what they like it or not, stuff like that. I mean, there were storage vendors still at the show. I mean, it wasn't like, I mean, IBM was probably the biggest, other than VMware, was probably the biggest booth on the expo floor, right? Yeah, that's different. Yeah, but we didn't see. Well, was Pure there?

I think Pure was there. Pure was there. NetApp was there. Tintry was there. HP wasn't there. A lot of companies weren't there. But, yeah, I mean, the big guys were there to a large extent. Yeah.

I'll tell you, that's one of the things, just real quick, that's one of the things that shocked me. Like, was like, wow, this is really different, is seeing the Dell booth. I was just like, that's the tiniest Dell booth ever. Yeah, I've ever seen a VMware. Yeah, and so they've, you know, they claimed like they stopped playing favorites with respect to ready nodes and things like that. So that, you know, there was an intense, I don't know what you'd call it, agreement between Dell and VMware with respect to, you know, VxRail and those sorts of solutions, which I guess to some extent still exists, but they're not the only one anymore. Right. Now everybody can play that game.

Yeah. And I can, I can see that happening, but well, let me go back to the storage part. So is VVOL still a thing? VVOL is still a thing. I heard VVOL 2.0 is coming out. But I think what it means is more HCL certification kinds of things. So they're going to start to certify the storage vendors with VVOL. And Tintree is still not playing the VVOL game, which I thought was kind of surprising.

Really? Yeah, they're not talking VVOL. And Tintree is still not playing the VVOL game, which I thought was kind of surprising. Really? Yeah, they're not talking VVOLs. And they never really adopted VVOLs to a large extent. They kind of went their own way. They've been doing a lot of the functionality of VVOLs without VVOLs. But yeah, the storage vendors I talked to, VVOLs is still a big thing. They're still going after it in a big way. And they think the VVAL certification will be a good thing when it comes out with VCF9.

So I'm just really confused about the way that Broadcom plays the game, right? So it's very, because they seem to have a very tight with everything like this is our plan and we just don't care this is our plan this is how we're going to be very focused on this plan and we're not gonna let um let anything persuade us to do anything different so it seems to me that the the storage piece of this seems a little weird to me that they're going to allow both. They're going to produce... They have been allowing

primary storage in their VCF solution for a while now. Even in the cloud. I mean, that happened to those guys. Yeah, I don't think they have a choice. I mean, people have... You really get to data gravity when you start talking about storage, right? Right right it's not an easy thing to do you know they're they're these storage solutions have their own set of features that people kind of grow to love and or hate and kind

of thing like that exactly that's very very true too yeah so what else did y'all hear the am i want to hear what else you heard at the analyst session. There was a lot of discussion on private AI as being, you know, the way forward for MLOps and AI functionality. I asked the question whether they ever saw training a large language model. They said probably not training, you know, 1,000 server, you know, 2,000, 8,000 GPUs, but maybe fine-tuning. But the one thing that I thought was kind of interesting is that, you know, they want to be. I think Private AI Foundation wants to be a Google Cloud Vertex on-prem, which I think is a pretty aggressive stance, especially since they don't really own an LLM, but I guess they can use any of the open source ones and things of that nature. Well, did they talk about, because what makes me think is, you know, where's the data and how are

we, how are the real data? A lot of it's for the big customers that which is their primary target right is inside of mainsframes so how many of their customer how many of their mainframe customers or vmware customers and is this a way to kind of spread that love around like you don't need to go to the cloud to do this stuff we have a private cloud that's better than anything right here. Yeah, yeah. I mean, they mentioned, probably it was the keynote, that they're seeing surveys that show 80% of enterprises want to repatriate their workloads from the cloud,

which I thought was a surprisingly large number. But then again, they're an on-prem solution. Yeah, but they're an expensive on-prem solution now. So that's the reason to repatriate is the expense, right? Well, I think, yeah, the reason, yeah, it is the expense of the cloud. And they believe they're much cheaper than the cloud for non-prem. And that's at least their perspective. Like I said, I don't know what the costs are

for licensing and that sort of stuff, but that's what they're playing the game. They want to be the on-prem cloud solution for all your workloads. The data problem for AI, it's multifaceted, as you know, Gina. It's not just mainframe players that have the data. I mean,

all these object stores out there and AWS and all these... There's a lot of data out there. They still have the data platform, data persistent platform, which is the object storage thing. They're still playing that game. So if you want an object store on VM, you can do this.

I don't know. It just, it is very expensive. So did they talk about, they didn't talk about any of that with the analysts i mean that's the thing i heard from the customers that i you know hung out with is just how expensive it is to the the prices are like 12 not 12 12 times increases when they get there i didn't hear anything like that from vmware and i'm not sure I would have in any case. But, you know, they certainly, it's a different focused VMware than it was in the past. You mentioned earlier before the call on the charitable and, you know, those sorts of organizations always had discounts for licenses. Is that the way it was?

Yeah, they always had education and charitable always had discounts for licenses. Is that the way it was? Yeah, they always had. Education and charitable always had discounts, and they don't anymore. Yeah, that's gone. I wrote a blog post. Yeah, I think I published it yesterday. I did some product marketing work on the market and how VMware has stated that they're going after it. But, you know, doing some research on that and hearing from a couple and reading a couple of case studies where, you know, that charities got their renewal costs just exorbitant. You know, it's definitely something they didn't think to plan for. And when they went into negotiations and, you know, took somebody with them to do the

negotiations, we're told that, well, VMware is just not for everybody anymore. So to me, it seems like, and the one example I saw was a children's hospital. That's what I wrote about and gave the link to. So, you know, what happened, what kind of ethical dilemma is that ethical dilemma? That's just business and everybody needs to deal with it. And I kind of find it hard because, you know, VMware, the reason it's, we all loved it is it's awesome. It does crazy, awesome stuff and it's solid. You can rely on it's awesome. It does crazy awesome stuff, and it's solid. You can rely on it.

You can rely on it to be up, and it's an integral part of any modern infrastructure. And so ripping that and replacing it with something else is worse than backups, right? It's hard. It's very hard. There's a lot that goes into it and a lot that has to be considered. And, you know, you think about, you think about some of these nonprofits, I'm not sure they have somebody running IT. Maybe they do, but, you know, maybe they don't get the people to run IT that think

about categorizing every application and figuring out how it's running and doing the economics and doing the, you know, like kind of managing that. That's stuff people kind of slip through the cracks sometimes because they're keeping the lights on every day. Right, right, right. How do you get to, I just, that to me is bothersome. It's great that they're doing all these things to solidify the product and give everybody a chance to use all of the new features

and creating a fabulous private AI platform. But who all's going to use it? I mean, they're focused obviously on enterprise environments and enterprise customers. And those are the guys that they think they want to tie their hat to and have done everything I can tell to do that. Right. Yeah. Let's talk about, you know, the subscription versus perpetual licensing transition as being part of a challenge. But I mean, there was a question about whether they had a different licensing for, you know, educational or charitable.

And the answer was no yeah as you're well aware yeah um but i just can't believe they didn't have any i i can see for other things but like no transition plan for these folks like okay we'll give you this much off the cost of the new cost now give you a year to get off of us since you can't afford us anymore. Yeah. I didn't see anything like that. Yeah, I don't think they have it. I think the biggest challenge with these sorts of acquisitions is probably less than 10% succeed in the long run, right?

Oh, really? Yeah, it's just that companies acquire other companies. They try to fold them into the product platform and things like that. The nice thing about Broadcom, they're really not that much in the software business, so this represents a whole new world for them. And I it's, so hopefully it'll, it'll succeed in the long run, whether it generates, you know, the billions of dollars they expect or not. I can't tell, but you know,

there's advantages to doing things like educational license and charitable license discounts and things like that from a learning perspective, from a just charitable perspective and things of that nature. Right? Exactly. But apparently, you know, the new world order doesn't seem like it thinks that's a much higher priority. Well, one thing I wanted to go to sessions to hear about, because I'm wondering if you heard about it, and I wasn't able to, was, okay, you've got, you said they talked a lot about the private AI stuff, and you could tell that because they had have, you know, they've got the CA, the security, which that feels like they are going to tie that together with a carbon black from VMware.

But maybe more from a chip perspective, do they have like some fancy new chip that's coming out that's going to power this private AI? Did they give any hints? I didn't think so. I mean, they're starting to support other GPUs. I think Intel Gaudi is one of the ones that they're supporting for. AMD was at the show, so I presume that AMD's GPU will be on that list sooner or later.

But I didn't hear them talk about any specific hardware chips. There was one session I was pretty, pretty vocal at. There was a discussion about whether, you know, Ethernet is the network of choice for AI. And, you know, there's been a lot of talk about whether InfiniBand, you know, is going to continue in that space or not. And this is the first time I've actually seen like a Broadcom hardware engineering manager there. Oh.

Talking about their Ethernet solutions and things of that nature. So it was pretty good. And they were pretty, I would say, vehement in their opinion that Ethernet is the right choice for backend GPU networks and things. That's interesting. That might set up another interesting technology showdown like happened with Ethernet in the beginning, right? Yeah. Yeah. I mean, the whole discussion about Finna band versus Ethernet is probably older than

both of us. Maybe not. Maybe not me. Maybe you. We can say that it is. I'm good. Okay. I like that idea. I don't know if I should say that.

But it was interesting. The whole private AI foundation is coming on a lot stronger than it was before. But to a large extent, it's the industry's movement into Gen AI has gotten such high publicity that everybody wants to be on this bandwagon. Right. But everybody will be on the bandwagon, I think. Oh yeah, for sure. I mean, how could you not be on the bandwagon? It's like saying you don't want to support SSDs 10 years ago or 20 years ago. You know, you got to, it's just, there's no solution here. Yeah. This is where we're going with, with computing. Man, that's, that's a lot. Like, that's kind of what I wanted to know is like, so what else are they, but you know,

now that you started talking to the other things now I'm interested in what from CA are they going to tie over to VMware? How are they going to tie the mainframes to a virtual backend? Nothing I've heard. They didn't even mention the word mainframe. I'm surprised you're talking about it, which is kind of interesting because the whole share thing was one of the Tech Field Day events here in the last week or so, right?

Well, last month. Last month, yeah. Time's going a little faster than that. Sorry, my bad. Maybe two months, actually. Now, we've been talking about, they didn't mention Watson X, which is, you know, the IBM AI solution of choice. Right.

I didn't see anything. But, yeah, because, I mean, that's kind of what well we heard it share you were you at share i don't remember no i was not sure i had another environment to go to but i see but the the thing that bmc and and um ca talked about broadcom um hang on one second i I'm sorry. I'm getting that time of the afternoon. Was a lot of it on how do you look of analysis on what's going on in there. I'm having a cough one second. I'm so sorry. That's okay.

Make a big drink of water. Maybe that'll work. They've talked a long time about COBOL as being, you know, something they want to move off of, but, you know, I still hear people looking for COBOL programmers, which is surprising hell on me because I've left that in the dust so many years back. I would never want to return to that. But, you know, this whole AI stuff, it's unclear why that, you know, if I was in IBM's shoes, I wasn't, you know, I would devote some amount of money to converting, you know, automatically converting or translating COBOL into some future language, call it Python or Rust. That's what they're doing. That's what they're able to do with generative AI.

And some of it's not converting. Some of it's just, because I mean, you know this better than me. I just know from i i took cobalt and assembly because they were my first two languages yeah so i but i know just from like the things we played around with how long the programs are i can imagine like a real production program like how many lines of cobalt is that but it's a bazillion lines it's a bazillion lines. And, and they were telling us that the, the programmers that are, are still going, because that the, the demographic of that like blew me away. The people at mom's age, people our age, a few younger people, but it's like definitely being gate kept by, you know, the original COBOL programmers that are still working. But they don't want to let any of the lines of code go. But those programs are not well maintained.

They're not built like a modern program. Exactly. Definitely not like a containerized program, right, where you have every discrete thing in a different section. They're not really marked up. They're not commented really well. So that's what I think kind of the first step is using generative AI to, and this was not, this was BMC, it wasn't CA. CA actually has this tool that'll go in and while it's analyzing all the

COBOL code, it draws the map of where the data flow go in and while it's analyzing all the COBOL code, it draws the map of where the data flow goes in and out. It will let them put comments. And just so you can kind of start cleaning up the code, but I'm sure the next step has to be using generative AI to modernize the apps. I was thinking more about the other thing just real quick that blew my mind is that data is being added constantly still to the main oh god yeah i mean these programs these programs just generate data to a large

extent as customers come in like i say vmware the analyst session the word mainframe was not even never even was uttered by anybody i should i should talk to you before so you could have brought it up. I should have. I should have. And ask my question. I mean, I know more about the mainframe than I care to admit, Gina. You'll ever forget.

Way more than I know anything about. But don't you, if they're building this, you know, and they're in it now for, because I was helping just kind of tie people together internally at vmware when i was there that were doing ai so they've been working on the technical bits for v-sphere and it's well documented how uh you can use you can use you can virtualize those workloads without a hit to performance which is what a lot of people especially in the hpc world were worried about they've been doing this for quite a long

time now but like how if the data if it if they're large customers and for the most part it does appear that they did the same type of acquisition with ca that they did with vmware you know they reduced sales and marketing costs and they concentrated on the very upper echelon of their customers. So if that happened, then these probably are the same customer set, right? There's got to be a Venn diagram with a pretty big middle, don't you think? I think to a large extent the big customers that CA deals with and the big customers that VMware deal with, I don't think are the same.

I mean, there might be a slice of that that are the same, but the guys at VMware are primarily people that work on Windows workstations and Linux servers and things of that nature. The people that CA works with is to a large extent mainframe users. Yeah. No, I'm not talking about individual users. I'm talking about companies. Look at the companies. I mean, if you look at, so I don't know, Visa, right? Visa or MasterCard.

Yeah, they got a big mainframe operation. But there's a whole huge segment of it, which is VMware only. Right. So like, what if those are my... You think there's a play here that maybe they could try to concentrate and convert, or at least, you know, join the CA customer base with the VMware customer base long time, long term. Or... There was not a breadth of that at the show. I know, but wait, this is what I'm trying to get to.

What if that's a way to keep those customers locked in? We've got everything you need from CA. We've got everything you need from VMware. We can help you figure out how to tie those environments together to get more for your buck. Yeah, maybe. Maybe that's a play in the long term. Play a book for Broadcom. It was not nothing, not even a hint, not even a breath. Next time you go, you got to ask. I guess. I should have asked. I'm just asking for Gina.

Yeah, I know. I know. I know. I mean, the whole VCF thing, to a large extent, is kind of a reimplementation of the mainframe because it does everything for you. It's your solution for containers. It's your solution for AI. It's your solution for workload automation. I mean, the things that you're talking about that they're coming out with or have been recently implemented, the mainframe, to a large extent, has had some of these capabilities forever.

That part's interesting, too, right? I guess. I don't want to talk about mainframe. You're bringing back bad memories. It's a bad side of my life. Oh, my gosh. I'm not sure I ever want to go back to. Like I said, I know more about the mainframe than I care to admit.

But back to VMware. Back to VMware. There's this whole bunch of advanced services that surround VCF, private AI being one of them. I'm just trying to figure out where they're going with this sort of thing.

I'm assuming private AI is a separately priced solution that runs on top of VCF, right? What's different about it, though? It's got NVIDIA partnership. It's NVIDIA focused, quite frankly. Well, so that means that you have to if it's you have to buy an nvidia certified vmware ready node kind of thing you know who's selling that nvidia selling it i think so that's

my understanding i mean you could do everything private you could do private ai with a reference architecture on standalone vcf with your own g, but you're not going to get all the functionality that NVIDIA brings to the table with their enterprise AI solution, which makes it vertex-like, really. Right, right. I think that would be, I would guess that you would probably have to have the correct license from VMware. They've only got like, what, four licenses now? So you'd have to have... Yeah, I think so. But I mean, this is like an advanced service that you deploy on top of ECF.

Yeah. So you'd have to have... That would be the base level architecture, but then you'd have to have whatever. I bet it's a... I bet it's not an extra... It can't be an extra thing. I think you would have to to have whatever. I bet it's a, it's a, I bet it's not an extra, it can't be an extra thing. I think you would have to pay for it. I think it is. I think it's a separately priced offering on top of VCF that if you want. From VMware or from? From VMware. If you want private AI, you get, you get this separately priced license for it or something like that,

that runs on top of VCF. this was a good question to ask chris oh yeah yeah yeah at the at the next session here are you going to ai field day because i think i am going to ai oh gina why aren't you going to ai field day that's another discussion uh i think because i did too much in august i need to i'm to have to pay myself back. You know, time-wise. You actually have to do some real work and stuff like that. I have to do some real work. So, I mean, back to these advanced services, they got advanced security, VDefend and web application firewall and malware prevention,

network detention response. All this stuff is packaged under this thing called advanced security. I mean, is VDefend, was that always part of the normal VMware solution? I'd never heard of VDefend before. So I don't know. That's different to me. So I wonder what that actually exists. So that's kind of why I wonder if this is just a packaging thing.

And it's because I thought they said you weren't going to pay any more than the one cost. That's why. I think you get everything with VCF. thought they said you weren't going to pay any more than the one cost that's why i think i think you get everything with vcf but if you want these advanced services you pay those cost more that's i don't know that for a fact but i believe that's the case uh they got the container operations you know they always had uh tkg and tanzu k Kubernetes Grid as part of the standard solution, but Tanzu brought all this other stuff with it. And Spring and all this stuff was all part of different packages

that you could deploy on top of this thing. So Tanzu application platform is yet another one of these solutions that is an advanced service that you could run on top of VCO. And they brought in all the spring stuff and Gemfire and all this other stuff. So it's kind of interesting. What is Gemfire? I don't know if I can talk about it. I went to a session. It's like a high-end data solution was Java-based and statistical analysis.

And, you know, I remember Qlik, was it Q-L-I-K, had this solution that ran a whole bunch of statistical analysis to try to understand, you know, your sales forecasting or something like that. And they would come out with the best match. But Gemfire is sort of a back-end solution for that. So they've got all this old-style machine learning, which really is more statistical analysis rather than deep learning kind of thing. And then they've got all this new AI, which is Python-based deep learning,

GPUs and stuff like that. And so they've got the two versions of this, which is a private AI and sort of the spring solution, which is now embedded under Tanzu. Yeah. You've got to ask about this at AI field day. Cause I won't be there. You've got to ask for me. Yeah. Yeah. Okay.

They've got, you know, they've got a whole bunch of, gosh, I don't know, disaster recovery. So what is VMware Live Recovery? I don't remember that either because that's something else. It's being able to recover your virtual machine solution to a cloud automatically, and it will try to maintain, you know, data synchronization and things of that nature. It's another one of these advanced services that you buy on top of VCF. There's a whole bunch of other ones. Load balancing, workload automation, edge orchestration, data service.

I'm looking at this slide that's got like 25, no, maybe 40 different line items on it. But that's all the different things that the BU's were creating individually and marketing individually. So are they just saying this is all the stuff and this is the place it goes into? This is where it runs. It runs on VCF. And if you want these solutions, you bring on these advanced solutions separately and package them. And they're packaged separately and apparently paid for separately, would be my guess. Well, that's kind of interesting. Yeah. So you're really just paying for the... Okay, I need some confirmation of that. Yeah.

That was good to get. I think I would require some confirmation. I suppose I could ask the right people there some confirmation of that. Yeah. That was good to get. I think I would require some confirmation. I suppose I could ask the right people there and stuff like that. Yeah. I might put some. So there were some enhancement to vSAN. They talked about replication now, asynchronous replication between vSAN clusters. They talked about deep snapshots.

What does deep snapshots mean? They've got 200 plus snapshots. Now you can have over a VMDK or a vSAN, I should say, data service, data solution. So you can now recover from recovery, you know, from ransomware and stuff a lot quicker. So it's more like storage now.

Yeah, it's getting way out. And that's what it was, right? The intent is to make it more and more like storage now yeah it's getting way out and that's what it was right i mean the intent is to make it more and more like a storage solution that that that you would go out and buy from primary storage yeah exactly that's interesting so yeah all of this is very interesting i mean i i guess that's the thing though you know it's just great but if only the very top echelon of enterprise can use it what what does that mean for everybody else yeah I mean they they still have this vvf which is you know the slimmed down version of you know what what was in before was

v-sphere and and its functionality so if you want just that, you can buy that. And it was like one of the customers at this breakfast I was at was a VVF customer. And they had, you know, they had primary storage and they were doing their thing. I mean, the other guys were fairly sizable companies. And they'll have VCF and primary storage. And we're looking at vSAN because, of course, they paid for it, you know i mean that's the thing that's what i find interesting yeah yeah is because i mean if you look at it that's the disruption now so like their their go-to market model changed so that's why there was some kerfuffle with the partners that i think is ongoing. Their pricing and packaging radically changed. And that's why you've got, you know, you've got some customers in some sticky situations

trying to make a decision. And they're even what percentage they're looking to get to when it comes to sales and marketing. That's changing. So that's going to have an impact on, you know, of course, what we hear and see and what's going on. So it's a radical change in the IT industry. Because if you're not able to stay or you don't want to stay now, that is not an easy, you know, it's not an easy migration like we talked about before. That's a big change. Yeah, yeah, yeah. Well, it's a new world.

It is a new world. It doesn't seem like they're going back from it. No. No, you know, that's their acquisition. And this has been their plan from, this is what they told us they were going to do from their acquisition. And they, this has been their plan from, this is what they told us they were going to do from the beginning. So, and they're doing it and they're doing it and they are executing on that for

sure. Following their playbook. I mean, you know, the analyst had a special session with the CEO and I thought that, you know, the guy seems very knowledgeable and has very good insights and stuff like that.

And he seems like he's trying to do the right thing from his perspective of where he wants the thing to go. And that's what they do. That's their job, right? Yeah, yeah. All right, Gina. Hey, this has been great. Any final thoughts you'd like to say on the conference?

Yeah, I was happy to see who I was able to see when I was there. It was definitely different. Drove home the point that it is a new world, as you say. And we'll see where it goes from here. Yeah, exactly. Well, thank you very much for being on our show today, Gina. Thank you for having me.

This was great. It was good. That's it for now. Bye, Gina. Bye. Until next time next time we will talk to the system storage technology person any questions you want us to ask please let us know and if you enjoy our podcast tell your friends about it please review us on apple podcast google play and spotify as this will help get the word out you