Broadcom Delivers the Modern Private Cloud, APJ | Exploring the Innovations of VMware Cloud Foundation 9.0

More from Broadcom Delivers the Modern Private Cloud, APJ

Paul Turner

Vice President of Products, VCF Division

Broadcom

play_circle_outline Empowering Customers with Seamless IaaS Integration: Delivering Tailored Cloud Services Through Comprehensive Compute, Storage, and Networking Solutions

play_circle_outline Broadcom invests 18% of revenue back into R&D, fostering platform innovation.

play_circle_outline Compliance and governance controls incorporated to manage data securely across environments.

play_circle_outline Integrating Kubernetes Runtime and Container Services into VCF for Rapid Deployment of Modern Application Stacks

play_circle_outline Cost-saving innovations like memory tiering and global deduplication reduce operational expenses.

Info
Transcript

Exploring the Innovations of VMware Cloud Foundation 9.0

Paul Turner

Vice President of Products, VCF Division Broadcom

Paul Turner, vice president of products, VMware Cloud Foundation, at Broadcom Inc., joins theCUBE’s John Furrier at the “Broadcom Delivers the Modern Private Cloud” event to explore the next evolution of VMware Cloud Foundation. Their conversation highlights the release of VCF 9 and how Broadcom’s integrated approach is reshaping private cloud strategy.

Turner outlines the journey to VCF 9, including Broadcom’s 18% R&D investment and its unified IaaS stack designed for consistency across on-prem and public cloud. With features such as fleet-wide mana... Read more

explore Keep Exploring

What recent advancements have been made in VMware Cloud Foundation, and how does it help customers in their cloud operations? add

What percentage of top-line revenue does Broadcom invest in research and development for innovation? add

What are the key storylines and features of VCF 9? add

What services are included in a platform that supports Kubernetes for running AI applications, and how does this address the complexity and training challenges associated with Kubernetes? add

What are the cost-saving benefits associated with the implementation of memory tiering and deduplication in VCF? add

bolt Powered by CUBE AI

Exploring the Innovations of VMware Cloud Foundation 9.0

search

>> Hello and welcome to theCUBE here in our Palo Alto Studios. I'm John Furrier, host of theCUBE for a special presentation with Broadcom VMware, delivering the modern private cloud with the launch of VCF 9. A big launch, really kind of coming off the acquisition. Now VCF taking the next generation next chapter. Here to break it down, Paul Turner, Vice President of Products at the VCF division at Broadcom. Paul, great to see you. Thanks for coming on and breaking down VCF 9. Everyone's been waiting for the news.

Paul Turner

>> Yes, John, great to be here.

>> Big journey getting here. I know there's been a lot of innovation, a lot of conversations, a lot of interest in VCF 9. Talk about the excitement of the journey. What got you here? What was the key focus? What did you optimize for? Break it down.

Paul Turner

>> Oh my gosh. Thank God we got it out. No, but I think the real exciting thing is we're really at a point where we can actually help customers make this next leap in innovation. Because 20 years ago we changed the world with kind of virtualization. What we've done over the last year and a half has actually integrated all of a full IaaS stack, compute, storage, networking, the automation that you need. You can deliver cloud on your terms. This means that you've got the same cloud operating model, whether you're running in your data center or on public cloud services. All of that is what we're delivering in VCF, in VMware Cloud Foundation. And in the 9.0 release, it's the culmination of that. So very exciting. The way we got here is we actually brought together every team inside in VMware and really got them to actually deliver on this integrated product. So it's a lot of work, a lot of IP, a lot of innovation in the product, and very exciting to get it out. It's been a long journey,

>> A lot of buzz inside with the team. Give an example of some things people excited about. What's the key?

Paul Turner

>> Yeah, so I think if you look at kind of key technology, one of them is how do we actually help an operator, an administrator, actually manage cloud at their terms and at their scale? And that means how do I do fleet management? How do I do updates? How do I make sure that patches are continuously maintained and non-disruptively applied onto my environment? So I've got a continuously secure, continuously trusted environment that gives me all the services, and it means then that the developer can run at their speed and run in an agile way.

>> Paul, really exciting on VCF 9. Congratulations. That's awesome. The innovation was a big story. We heard that from last year. What's the innovation coming out? Can you share some of the points around the innovation around VCF 9? What should we look at?

Paul Turner

>> Oh, yeah, it's a really exciting release. But you think about innovation, I'd like people to also think about innovation and Broadcom because Broadcom actually invests 18% R&D of top line revenue back into R&D innovation. Now compared to other companies, you can go back and look at stats for somebody like Microsoft and you'll see it running at around 12.5%. So where do we put that innovation? We put that into research and development. We put that into building product. VCF 9 is the culmination of that innovation. So what we've done inside that product is not just delivered a platform for customers to deliver cloud on their terms. Remember, many years ago, we changed virtualization 20 years ago, but virtualization isn't enough. Developers want agility. They want software-defined compute storage network, the automation associated with it. That's delivered in the modern private cloud, and that's delivered in VCF 9. So it's very exciting by lots and lots of engineer hours and many, many months of work.

>> That was the promise, the innovation. You guys bring it to the table. VCF 9, it's the crown jewel for VMware and Broadcom Software group. What's the key story lines? Take us through VCF 9.

Paul Turner

>> Yeah, so I think if you think about this. This cloud operating model, what are we really trying to do? The data center and the way people run operations, it has policies, it has governance controls. It's how do you actually manage and how do you deliver them for developers, a speedy, agile cloud experience, that's what VCF 9 is. So inside in VCF 9, we've actually got a whole fleet management layer to actually help the administrators, the operators of the platform actually run the platform at scale and deliver it cost efficiently for our organization. So capacity scaling, reducing, looking at inefficiencies, looking at green data center, looking at compliance rule sets, all of that. So they can keep the governance rules, they can run it anywhere, but the governance rules they can enforce everywhere. And then of course, it's all about developers. How can we help developers move applications quicker from instantiation and ideation of it, right through git level management right into production operations. That means the platform needs to act like a cloud platform and deliver services on demand. And that's the other aspect that we deliver is full application platform, virtual machines, containers, AI, ML applications, all of that in an agile way for developers.

>> And the platform engineering side is setting the table for the gen AI movement coming in. That's going to be a big part of it.

Paul Turner

>> Yes.

>> The Kubernetes has been stabilized. People are rallying around that. You're starting to see that stability moving in and setting the table for the gen AI, which is chaotic right now, but it's starting to normalize around the clear value in enterprises. On-prem action for AI is hot and then you got sovereign cloud, you got all these other things going on. This is the private cloud moment for VCF.

Paul Turner

>> I think it is the private cloud moment. And yeah, you do think about AI applications, right? AI applications are actually built on container-based services. And those container-based services means you have to deliver a platform, which is a developer ready platform that actually has VM's and containers on the platform, check included. That it has the automation that the developers need so that they can actually get access to services on their terms. They don't need to go through a queuing operation into IT to request services. They can get them on demand. Check. We've got that included as well. One interface for the developers. The admins, as I said, have all the controls that they need, but the last is it needs that security, control, that sovereignty of data, that ability to kind of manage a secure and trusted cloud operation wherever it runs is the third aspect. But yeah, you've got to build a platform that is a cloud platform and it's a private cloud platform.

>> It's super exciting. The scale, the scale, the data, the collision between platform engineering and analytics is coming together. And that integration shift is clearly on the table. We're seeing it in the stories we're covering. It's a big part of VCF 9 integration.

Paul Turner

>> Yes.

>> It's clear. What shifts are happening that you guys are integrating into and some features of VCF 9 that you guys are taking advantage of because this is really relevant right now?

Paul Turner

>> Yeah. Well, I think of the admins. We've got 300,000 plus customers. We have 120,000 people who are part of VMUG members. Those people have actually helped change IT, change the world of IT. We're actually helping those customers not just change the world of virtualized compute and the ability for us to standardize IT with virtual machines, but actually helping them run the entire cloud platform and actually to virtualize networking, virtualize storage. But how are we helping them? They get single centralized fleet management. They get config management, so they can do compliance and configuration compliance across their infrastructure. They get certificate management so they can actually monitor and confirm any of the certificates that are out of date or non-compliant. They get the capabilities for doing diagnostics where we look at common issues that we've seen in other customers. They get security alerts in context based on what we see as common threats and issues that they need to be aware of. So the developers, sorry, the admin's life is really made much easier. How can we help them scale? How can we help them deliver a cloud platform. A cloud platform, not just a virtualization platform, but a full cloud platform for them. And then the other aspect, of course, is as I said, the developer productivity. The developer productivity where we give that consumption experience to the developer, the speed that developers can write and deliver kind of code level, automated code pipelines, CICD pipelines where they deliver it directly against infrastructure. All of that's possible. And of course, containers, VM's, AI applications.

>> Integration as a feature really seems to be one of the highlights here at VCF 9. When you guys are looking at the product roadmap, I mean that comes from customers, that comes from the market. Is that a key takeaway?

Paul Turner

>> It is a key takeaway, and it's one of those takeaways that you look back and you say, "Gosh, we should have done it many years ago." I mean, VMware had the best assets for delivering a private cloud operation. We have the leading software-defined compute with vSphere, the leading software-defined storage with vSAN, the leading software-defined networking with NSX, but the integration that we had done between those products, the think about API's where the API should have the same kind of deprecation. How long do you keep them around for support terms and durations. SDK's and developer kits that is all unified in this platform really? So that makes a big difference because it just reduces complexity. And complexity is a big thing. I look at Kubernetes space. Kubernetes is a complex area and it shouldn't be. And what we've done on VCF, we can bring to Kubernetes and we are bringing to Kubernetes as well, because complexity kills administrators, it kills scale, it kills fleet management at scale. So you've got to reduce complexity. What we did is we converged all the teams together. We got them to work on one product over the last year and a half, and we're delivering it now with VCF.

>> And that's great and that's great, and that's where the innovation comes in. I want to zoom out and look at, there's a lot going on. Break down the key highlights and the pillars of how to think about VCF 9. Okay, there's a lot of value there. You touched on some of those key things. How should we frame this? Is there a framework? Are there pillars to think about? Lay out the thinking around VCF 9.

Paul Turner

>> Yeah, well, I think the easiest way to think about is think about how do I help, it's always about how do we help make people more productive? How do I help an admin team be more productive to deliver a cloud operating model for at fleet level scale? And how do I deliver for developers an agile platform so that they can deliver the time to get applications from ideation through production is very rapid and quick and on their terms, and then security and compliance. So those are the kinds of key things. But I think one of the most exciting things I think about beyond the integration work, we actually have also been driving innovation into the platform. And one of the exciting things is the work that we're doing on modern apps.

>> How do you guys stack up against the public cloud because you mentioned earlier, this is private cloud. We see the benefits obviously with distributed computing. For AI modern apps this is an alternative. You mentioned sovereign cloud as one example that's booming, going to see a lot of action coming on rest of the year and going into 2026. How do you compare against public cloud? Everyone's going to be like, "Okay, why do you compare against the cloud?" What's your answer to that?

Paul Turner

>> Yeah. Well, you think about kind of Kubernetes and runtime AI applications are container based applications. That means you need to have a platform that actually runs Kubernetes incredibly well and actually gives you all the services that you need. And you think about the service platforms that can do that. You've AWS EKS, you've got Google GKE, and you've got VMware VKS, which is our vSphere Kubernetes service. So natively built into the platform, we deliver a full Kubernetes runtime. And it's really important. If you think about a survey that CNCF did on this, and this was actually their 2023 survey, but 46% of people said that the problem is the complexity and lack of training is the problem with Kubernetes. And why? Because we're getting our most expensive resources, our developers to actually go and maintain the platform services. So it shouldn't be the case. So what are we doing? We're actually making sure that every service that you need to actually run is there part of the platform. So you get a Kubernetes runtime and multi-versions of it. You get a service mesh, an Istio service mesh. You get a Pinniped integration into single identity. You get Prometheus for monitoring. You get Harbor, the registry service that you need for container registry. You get integration in with your storage and your networking with your CNS and CNI adapters. You get your multi-cluster management so that you can actually manage across multiple regions. You get Valero, a backup service so that you can actually protect all of your applications. All of those services previously developers would have to manage and maintain. They don't have to with VCF. They're all part of the service, part of the platform delivered for them on demand. They can just write applications, work out how do they actually go from code idea to code delivery is much quicker.

>> The app dev space is hot for this. There's a great, great call out there. Let's talk about the customers. What's in it for them? What are the reasons they're going to get excited for VCF 9? What's the data show? What's their reaction going to be?

Paul Turner

>> So well, I hope they'll be thrilled with it and I'm sure they'll be thrilled with it. But often the way you look at, let's kind of think about stats and ROI kind of benefits from it. And one of the things I like to look at is, is it going to save cost? And we've put in memory tiering the ability for us to do something pretty cool. Nobody else can do this. Take DDR 5 memory, take NVMe storage, magically merge those into one pooled entity. Without impact to the applications, you can lower your server costs by about 38%. That's pretty cool.

>> That's huge.

Paul Turner

>> With storage, I get a 34% reduction because we've actually added in global dedupe, not just dedupe on a single deduplication on a single server, but across an entire cluster-wide deduplication. So 34% lower cost in storage. So VCF isn't expensive. VCF is about saving money for customers and bringing .

>> The memory tiering you bring us. That's actually a great example of some of the innovation. Memory tiering has massive impact on costs and customers will see that directly in VCF 9.

Paul Turner

>> It's too expensive of a resource for people. 50% of the cost of the average virtualized server and compute server is memory. But we didn't just stop there because those are all about cost savings. The other thing you got to look at is can I deliver a better available continuous uptime and efficient platform? Something that we've done forever is we know how to take virtual machines and applications, virtualization of applications and squeeze them into more resources. Imagine we could do that for AI applications. Well, we can. So we can do it because we've taken our vMotion and we can do totally non-disruptive GPU vMotion in 9.0 where we can take running workloads that are AI workloads, running PyTorch, running Lama II models. This was the testing we did. Lama II models, 70 billion parameter models. We take that server out, we move the workload. Everything magically keeps running. Your inferencing still runs, all of your AI applications still run. Nobody else can do that.

>> Scope the benefit there. I think that's important to kind of highlight that. That feature. What does that mean for the customer? That means no downtime, configurations. Workload management.

Paul Turner

>> No downtime. It allows them to actually get more resource usage out of their current GPU's. GPUs are expensive, 160 gig of memory sitting there. And so you've got an application that's going to consume 80 gig of it. I've got 80 gig left over. What happens when my next application comes in? Can I fit it onto this? I've got a pool of GPU's. I want to efficiently manage them. DRS for us does automatic placement of them. If we find we can't place, we can rebalance AI workloads, non-disruptively again, or maintenance situations, drivers, hardware. Things need to be updated on physical hardware sometimes. So it's a huge gain.

>> It's interesting, Paul, is that how many times have we heard in this century outcome focused/ outcomes are workloads because end to end, you have full visibility on the workload managing the workload. The outcomes now are clear. You can actually work backwards from the outcome. It's not a cliché anymore. It's like legit outcome. Why is that important in VCF 9? This seems to be a big benefit.

Paul Turner

>> Well, I think for any of our customers, our customers are always looking at TCO and the ROI from anything that they do. At the end of the day, they wouldn't buy our software otherwise. Our software is about saving money, giving them agility, giving them better availability of the platform. And oh, by the way, we can do it with no overhead. So I love the work that we've actually been doing. We've actually been improving virtualization right down at the core. So you can actually get a full virtualized environment versus bare metal with about one percent overhead. And that's cool.

>> As you're having the keys to the kingdom as VP of Product at Broadcom VMware, you have to look at the holistic picture of what customers are doing, what you're doing internally. You've simplified internally, you're talking to customers that input into VCF 9. A lot's going on. You guys have been super busy. As you go forward, this is the Broadcom innovation we kind of touched upon at the top. Broadcom's playbook is innovate on behalf of the customers. So every time in all aspects of Broadcom, they add more stuff. What's next? What are you guys working on as Broadcom VMware Explorer comes up like a few months. I'm sure there's going to be a lot of action. Can you share a little bit, tease us about what's coming, what's on top of this? Obviously this rolls out, customer will get it, it's an upgrade. What's next, what are you peddling around?

Paul Turner

>> We're continuously innovating and that's kind of the exciting part. VMware has not changed. VMware by Broadcom hasn't changed. We're a font of kind of innovation and that's why you put 18% back of top line revenue back into R&D. So you're going to see improvements in terms of our DRS and vMotion pieces. Further work that we're doing there to actually allow kind of full dynamic encryption level encryption capability so that we can start moving workloads around in an encrypted manner, decrypt them on the fly and do that very dynamically and without impact again, to the applications. You're going to see us invest in kind of core technologies. We've got to look at our vSAN and our storage side of things. We're going to look at kind of new media types that we can support so we can actually support a much more dense and deep storage, something we call cyber store with full protection. So you don't just get an ability to run an application, but you're getting full disaster recovery and you're getting the ability to rewind time. If you get infected by, and this is possible today in VCF 9, and we're making it actually possible for on-prem customers to do this, but they can rewind time and undo any of the damage that may have been done by some cybersecurity threat into their environment. So that's cool. And then of course, I can't ignore what's happening in the geopolitical climate right now. We're seeing a huge push from our customers and more big investment from us in sovereign cloud and sovereign AI. The ability for you to have built on our private AI work that we have done in private AI foundation, but how can we bring that together with sovereign controls, sovereign cloud, with our CSP partners, our cloud service provider partners, because we don't just deliver into data centers. We deliver to 500 plus cloud service provider partners. We deliver to all the major cloud vendors, whether it's AWS, Google, Azure, but the sovereign cloud is particularly interesting, I think, for their CSP areas. So that's another area of innovation. And of course we're going to continue to innovate on modern apps. It's great when you have thousands and thousands of developers, which we have. The level of innovation you get is phenomenal.

>> And the customer base, you've got input coming from customers. It's critical infrastructure. Private cloud is powering AI, it's powering distributed computing across the world and sovereign cloud, we see that massive opportunity on country, whatever word you want to call it. There's need. Private cloud is key infrastructure even in the data center.

Paul Turner

>> Yeah, absolutely. Because the big training of the models will of course be done on massive infrastructure, 500, 600 GPU infrastructure delivered from the Meta or Facebooks of the world. But because of the work that they did, it allows us to actually do cloud for businesses. Can take a lot of those learning models, run them in a private, secure, trusted way, run it on sovereign data. That means sovereign soil run by companies, your company, which is local within the particular region and is a trusted entity within that region and deliver AI services on top by really building, and this is the cool thing about gen AI. You're building off the bootstrapping from the foundation of others.

>> Yeah, I mean it literally is at the foundation, the IaaS layer and the software layer that's going to have enable all that stuff on top. Congratulations on VCF 9. It's out there. Big milestone. The innovation continues. If you had to kind of put a bumper sticker on VCF 9, Paul, what would it be? Would be like cloud for everyone? What would be your highlight?

Paul Turner

>> I would say it's basically it's cloud on your terms because it delivers you that private cloud, that operating model, that ability for a company. And we deal with the biggest companies in the world. How do they put their governance rules, their compliance rules, whether it's a finance industry, a healthcare industry or anything. How do they make sure that the platform delivers that and they know that they're compliant and they can deliver it anywhere? And that is very powerful and it's only possible when you virtualize everything with software. And that's what we're doing. We're giving you the ability to virtualize with software and you kind of abstract yourself from any custom cloud model because AWS is a custom cloud model. Google is a custom cloud model. Azure is a custom cloud model. Not with us. When they work with us, you get that common cloud operating model.

>> Yeah, customers are building the bridge to the future. The VMware has been there in all the top accounts in the network. Congratulations. Thanks for coming on, sharing.

Paul Turner

>> Thank you.

>> VCF 9 here at theCUBE. I'm John Furrier here for theCUBE and Palo Alto students for special presentation with Broadcom VMware. Delivering the modern private cloud with the launch of VCF 9. Thanks for watching.