VMware Explore 2024 | Mike Gannon | Broadcom & Stephen Elliot | IDC

Clips
News
More from VMware Explore 2024

Michael Gannon

Chief Revenue Officer

Snowflake

Stephen Elliot

Program Vice President, Management Software and DevOps

IDC

Reorganizing cloud strategies: A new era for private and hybrid clouds

A decade ago, public cloud was the de-facto standard, promising scalability, flexibility and cost savings. Given the explosion in data volumes and proliferation of workload types, companies are now reassessing cloud strategies and moving select workloads to private and hybrid clouds.What’s the root cause of this seismic shift,and how will it impact artificial intelligence operations?Discussing cloud strategy with Broadcom’s Mike Gannon and IDC’s Stephen Elliot.“We saw that there was some labor shortages in the lift and requirements to get some traditional applications there,” said Mike Gannon (pictured, left), president of North America at Broadcom Inc

play_circle_outline Emerging Trends in North America: Importance of Workload Rebalancing in Private and Hybrid Cloud Environments

play_circle_outline Emphasis on cost, resiliency, growth opportunities, and driving efficiencies

play_circle_outline Value of private cloud over public cloud for certain workloads

play_circle_outline Importance of reorganizing silos into platform operating model

play_circle_outline Focus on product innovation, sales execution, customer empowerment for success

Info
Transcript

Mike Gannon | Broadcom & Stephen Elliot | IDC

Michael Gannon

Chief Revenue Officer Snowflake

Stephen Elliot

Program Vice President, Management Software and DevOps IDC

Mike Gannon, president of North America at Broadcom, and Stephen Elliot, group vice president for infrastructure and operations at International Data Corporation, discussed the shift back to private cloud and the importance of optimizing workloads for efficiency. They highlighted the importance of reorganizing teams to support a platform operating model and mentioned the cost advantages of owning a fully integrated private cloud stack like VCF compared to a public cloud experience.

They also emphasized the need for strong executive support and cultur... Read more

explore Keep Exploring

What are some of the reasons for the shift away from moving all workloads to public cloud and towards building private cloud infrastructure? add

What factors are driving CIOs and CEOs to reevaluate their technology architectures and consider moving workloads back onto private cloud repatriation? add

What factors are driving the decision-making process in leveraging technology architectures for businesses, particularly in terms of potentially moving workloads back onto private cloud? add

What challenges do developers face when bringing workloads to a native instance of a hyperscaler? add

What are some key strategies that Broadcom is implementing in order to drive success moving forward? add

bolt Powered by CUBE AI

Mike Gannon | Broadcom & Stephen Elliot | IDC

search

Dave Vellante

>> Hey, everybody. Welcome back to VMware Explore 2024. You're watching theCUBE. My name is Dave Vellante. I'm here with my co-host, Rob Strechay. This is day three for us, extracting the signal from the noise like theCUBE always does. We're super excited. We're going to dig deep into infrastructure and operations. Mike Gannon is here. He's the president of North America at Broadcom, and Stephen Elliot is going to bring the analyst perspective. He's a group vice president for infrastructure and operations for the cloud operations group and DevOps at International Data Corporation, my old company. Great to have you guys on theCUBE. Mike, Stephen, welcome.

Michael Gannon

>> Thanks for having us.

Stephen Elliot

>> Absolutely.

Dave Vellante

>> All right. Mike, let's start with you. What's the big picture? What are the trends you're seeing in North America? What's the vibe out there?

Michael Gannon

>> Well, I'll take the vibe from what we're seeing at the conference, which is, from the practitioner level, a ton of excitement around what we're doing around building private cloud. I think if you look back 10 years, massive interest in moving as many workloads to public cloud as possible. We saw that there was some labor shortages in the lift and requirements to get some traditional applications there. We started recognizing some cost concerns around once the application was there, was it the most efficient place to run it? So we're starting to see a rebalancing of what is the appropriate workload to run at the cloud. And I think from the IT practitioners that are participating in these conferences, they're excited to start getting back into the game of helping create a cloud operating model that, whether it's depending on a private cloud instance or a hybrid cloud instance. The VMware community was vibrant here today. They're excited at this VCF messaging. They love the ability to start delivering public cloud experiences on-prem, and then we start providing them some really neat ways to have workload portability between Amazon, Microsoft, Google. That's an exciting concept because the IT practitioners who've optimized applications on current data centers are now part of the cloud experience. Whereas in the past, it was the developer ran off and they wanted to start optimizing the app, but the IT wasn't really involved in the decision making. So we're seeing this repatriation of not just workloads but the talent pool that's now supporting this cloud experience.

Dave Vellante

>> I have a premise on this. I wonder if I can ask the analyst's perspective. 2008 in the financial crisis, you had an uptick in public cloud because the CFOs were saying, "Hey, we don't want to burn the CapEx. Let's just rent the infrastructure," and that was a way to sort of preserve cash. Cash was king. And then we come out of that 2010, let's call it, and you had zero interest rates for a decade. Nobody seemed to care about cost, spend, spend, spend, and then you saw that through COVID. "Oh, we got to go to the cloud because everything's remote." And then all of a sudden overnight, it flipped where people said, "Wow, cost actually matters," looking at their cloud bills. And then to Mike's earlier point, it's been kind of a rebalancing then. how do you see it, Stephen?

Stephen Elliot

>> Well, there's a lot of CIOs and CEOs looking at costs. They're looking at efficiencies and I would say now you look at resiliency for their business. You look at growth opportunities. You look at how you can drive profitable growth. The recognition on where to place workloads, how to think about what these workloads mean to their business has elevated at a level we haven't seen before. One of the reasons is you look at COVI and how that's driven the decision making relative to how do we leverage the technology architectures to drive our business. We're looking at opportunities relative to moving workloads back onto private cloud repatriation. In fact, we did some research in 2023. We spoke with 2,250 technology executives and we asked, "Are you planning to repatriate workloads in the next 12 months?" 80% or more across compute and storage said, "Yes, we are." And I think as they're starting to get a sense of the compliance situation, the governance, the security, even gen AI this year, you're seeing more and more CEOs say, "Look, let's make the best decision for the workloads. Let's take advantage of the skills we have bought..." Not bought. "The skills we've developed over the past decade. Let's look at the importance of private cloud. Let's think about how we can drive efficiencies through process automation." All these key requirements are really re-engaging relative to CEOs and CIOs but also re-thinking, "Let's make the best choice." And that's why a lot of these executives are saying, "Hey, let's pull a lot of these workloads back onto private. Let's drive the efficiencies we developed. We have control, we have governance, we've got line of sight and the cost. And with gen AI, we're making decisions about models, large language, small language. We're making decisions about data, where the data is, who owns it, who has access." A lot of these themes represent some of the best of what private cloud offers and I think that's why you're seeing this engagement. Finally, I'll say, look, public cloud, there's a use for it. It's always going to be here. We're in a multi-cloud world. But now as executives think about the security, think about the governance, think about the complexity of who has access to data, the level of risk they're willing to take, we're seeing executives say, "Private cloud is a better option for a lot of workloads. Let's make the move."

Dave Vellante

>> That aligns with Hawk's three Cs: cost, complexity, compliance or control as Stephen just said. But the narrative from a lot of folks is, "Oh, well, Broadcom VMware's raising prices." Help us understand why your new model is... It fits in to what Stephen was just saying and drives better TCO.

Michael Gannon

>> It's a great question. When you look at what we bring to market now with VCF, fully integrated private cloud stack, we actually took the price of that and cut it in half. We made the cost of entry lower by 50%. Selectively, you look at what it costs now to run a private cloud on VCF and you start comparing that to a perhaps public cloud experience. We're a quarter of the cost. Virtual CPU on-prem, $104 per virtual CPU. This is... By the way, Alan Davidson, our CIO is the next target for you to speak to. He did a massive engagement around the cost of running IT. $104 per virtual CPU per year and he's got a massive investment in the hyperscalers, $495 per virtual CPU. So you look at how we've positioned and priced out VCF for general IT consumption, it's still incredibly well-priced as against to a comparative public cloud consumption. But then you start thinking about the OpEx and CapEx elimination when we start driving virtualization deeper into an enterprise and we start to virtualize more of compute and networking and automation and operations. The economy is the scale you get out of this and we've got customers running 3 million cores of VCF in production, mission-critical workloads. We think it's absolutely perfectly priced for the market. Is it an adjustment from what it used to cost to maintain a vSphere kernel? Yes, it's an adjustment, but it's a full stack. So we wanted to compare ourselves from a price point to a public cloud experience delivering that on-prem.

Rob Strechay

>> But to Stephen's point about bringing... And I think you even said it as well, bringing and recapturing that knowledge and those people and that staffing to be able to get at that public cloud-like experience on-prem. And we totally agree that public cloud operating model is not a place. It is a model and an operating model.

Michael Gannon

>> Oh, absolutely.

Rob Strechay

>> How do you see that happening? Because you're out talking within North America, obviously, the licensing. Some people are up in arms and all of that. We'll put that to the side. Beyond that, they also have to look at it... And I was with one of the customers yesterday having a discussion and his whole thing was, "I'm now bringing together my entire stack of VCF and I'm bought in and I have that stack here." But what I had to do is I had to bring along the guys who were doing Cisco and Arista to now bring on NSX so that I can see that value. What are those conversations going? Because that's new to some of that discussion as well.

Michael Gannon

>> Two-part answer. I think it's a wonderful question. For developers that started bringing workloads to a native instance of a hyperscaler name your flavor, they went there particularly for the microservices that hyperscaler can provide to the application. That developer soon found themselves spending 60% of their resource time doing all the things that traditional IT used to do for them: the governance, compliance, backup recovery, performance managing, resource management. Only 40% of their time was on building business logic to modernize the application, so a new complex silo started to evolve based on workloads that were being developed natively. Now you start bringing it back to the practitioners who ran a highly virtualized data center saying, "Hey, I want to give you that same experience and I can deliver that to you in public cloud with our partnerships with Amazon and Google and Microsoft. We give you that VMware instance, so we give you that workload portability, a seamless portability from on-prem to off-prem." So we give you that consistent control plane, that consistent developer experience now that developer's focusing on business logic. But when you talk about how you're going to build an operating model support private cloud, we have to break down some silos. The storage team is only responsible for their services that they provide to the app team, ticket gets handed over to the network team, they have to worry about advanced load balancing, firewalling, security measures, but they're working and operating in silos. So the operating model that all of our customers that have deployed VCF have done, it's been a tops down. We have to organize ourselves differently to a platform operating model versus the discrete silos. VMware had to do this from an R&D perspective. We built great products. We had five business leads that developed wonderful products. In order to make VCF the platform of choice for private cloud, we had to reorganize ourselves. So we've got one gentleman, Krish Prasad, that owns the entire stack so that I don't have storage teams running rogue on doing something just for the storage market without doing full integration and lifecycle management against VCF, so we've had to reorganize ourselves to deliver the experience to the market. Customers are finding themselves reorganizing their silos to have platform teams. Just like a developer would have platform engineering, we want an operating model that has a team that's delivering a cloud operating model, not just discrete storage network and compute services.

Dave Vellante

>> This has been a long time in coming. I mean, I go back to 2009, 2010 when people were talking about private cloud to private cloud experience, trying to replicate to a substantial degree the public cloud experience. Hock Tan yesterday basically said, "You get AWS on-prem." Now at the analyst meeting, he wasn't as definitive about that, but he's like, "Look, it's basically very similar, substantially similar, let's call it." Assuming that's the premise and assuming if you go all in on VCF and by the full stack, you're going to get a cost advantage of owning versus renting. We can all agree on that for now. What, Stephen, do you see from an analyst perspective that Broadcom VMware has to do to sort of maintain that, given what you know about the pace of cloud innovation?

Stephen Elliot

>> I think there's a couple of things, some of which is already occurring or has occurred. Number one is focus on product innovation. It's something that Broadcom as a company has had historical legacy excellence at. VMware has also had that historical execution, so product innovation relative to... You look at VCF 9 coming out. You look at the role of automation in that. You look at the role of integration across some of the product suites. You look at the addition and some uplifts on security. These are things that we're expecting to continue relative to driving, making the private cloud like an AWS experience. I think the second piece is, as a company, they've organized around four business units. The focus for each business unit is very specific and there's a lot of accountability, so that's another piece relative to ability to execute. A third piece, I would say, is working with customers. Working with customers, especially top 1,500, 2,000 accounts, and really empowering them with an understanding of, "I'm moving from a product perspective to a platform perspective." Broadcom is looking at value tools to help them understand, "Here's the value we've built for you today. Here's the value that we're going to deliver for you as a platform." And besides that, also the ability to deliver prescriptive guidance, because this is not easy for technology teams. So breaking down those silos, as Mike just mentioned, and helping them understand, "Look, we've delivered great business value for you and here's what that means technically and here's what it means from a business perspective: be the translation layer. And now as we move to new areas of expansion such as networking or security, et cetera, here's the additional value we're going to drive for you and let's make it easy to make that transition." If they're happen to do a tool replacement, de-risk that migration, provide services. The other piece is, and I think this has been very clear, there's a long tail of customers here. The partners are going to play a really important role in the success moving forward. That's clarified more than ever. You have four different types of route to market partners. You have four leaders at Broadcom who are very specific about the changes they've made and execution moving forward and how Broadcom is going to work with each partner to drive profitable growth. When you put these pieces together, you start to see it's not just having a great product. That's no longer good enough in the technology industry. You've got to have really great sales execution. You have to rethink the relationship and then empowering customers to say, "We understand where we are today. We understand the value and this is where we want to go, so help us with the roadmap. Help us with prescriptive guidance, whether it's direct or through a partner, and then take us through this journey." And I think those are the types of changes that we're already started to see and we're going to continue to see and we expect to accelerate over the future.

Rob Strechay

>> How do you see the customer conversations that you've been having this week going? Because obviously some are all the way through their transition in licensing to the BCF platform. Some are still trying to figure it out. We have some anecdotal evidence of others that are seeing what's coming out of this week to make their choices. What are you telling them about that? Because obviously there's other platforms they can go to. They could go all cloud, they could go to another hypervisor, or they could limit their use to, "Hey, this is my core VMware and I'm going to move other stuff other places."

Stephen Elliot

>> Yeah. I could put our customers into one of four stages of the transformation stage, stage one being, "Do we have vision alignment around what your cloud operating model aspirations are? Do you want to be private? Do you want to be public? Do you want to be hybrid?" We have to identify what their cloud operating choice is. The second item, and Steven spoke about a little this before, is we do value modeling. Let's just take a look at how you're running IT today from a cost perspective. Finance is the language of the business. Let's identify what it costs to run IT today and we get great telemetry out of the VMware state so we can identify, "Here's your current state cost model," and then we show them what a future state cost model would be with a highly integrated automated BCF stack. So we give them the financial ROI and TCO and the payback period on what this fully deployed would look like. The third and probably most critical I would say, and this is where I've seen some failures in cloud deployments, the consumption plan. How are we going to get you from current state to future state? It's not going to all happen day one. It's going to happen over a period of time through the term of the contract with our professional services teams either delivered directly from my organization or through our vibrant partner community. We're going to talk about how we start taking refresh cycles and taking that refresh cycle of compute storage and networking and integrating that into the stack, so we have to deliver that consumption and delivery model. And then the last but not least is when you get into the actual contracting on how we're going to help ease you into this from a costing perspective over a three or five-year term with stepped up laddering and invoicing. we've got one of four stages and the last stage is we've exited the contract and now boots are on the ground delivering transformation. And we're doing quarterly check-ins with the CIO, "And here's where we were three months ago. Here's where we are now. We're delivering value. Here's where I need some help. We've got a little bit of bureaucracy preventing us to get certain elements of the stack deployed. Need some tops down support. That's going incredibly well.

Dave Vellante

>> Number three, obviously, is the critical mandate, getting that consumption going-

Stephen Elliot

>> Absolutely. ...

Dave Vellante

>> getting the adoption of the four SKUs and the micro SKUs and particularly VCF.

Michael Gannon

>> Yeah.

Dave Vellante

>> So I'm interested in some of the things that you guys have done to accelerate that. I mean, one being... And even in 5.2, I think, Rob, you were sharing with me that the migration from vSphere to VCF used to be really painful. Now it's frictionless because you've done the engineering work. Hock sort of said in our session, "Got to do the hard work. That's an example."

Michael Gannon

>> Yeah.

Dave Vellante

>> What else are you guys doing to make these transitions easier and specifically the consumption to get to that North star?

Michael Gannon

>> Yeah, I'll go back to... I think the technology is the easy part here. It's our people and process modernization that really is where I need a cultural tops down partnership between the CIO and CEO who's got his back to say, "Hey, we're going to break some things along the way, right?"The relationship of the CIO and the CEO is incredibly important. "If we're going to go and be successful at this, we're going to break some things but I need your support when people start hollering at the business to say, 'Hey, I got your back. We're driving transformation. It's not easy.'" Spending more time on the process of which how a developer comes in looking for a service and how we're going to automate that ticketing process... Process is critical, right? The silos, breaking down the procedural elements of how you hand off all these services through the supply chain of It and saying, "Hey, we got to break down these silos. We're going to do things differently." That, to me, is where we need the most amount of support from the executives of our customers to say, "Hey, I'm going to help break down some cultural barriers for you so you're successful." Because we recognize with AI coming, it's a tsunami of new data and governance and complexity. The infrastructure has got to be a hard bedrock. It's got to be absolute granite. If it's not granite today and all this new business demand comes at you, everything's going to fall apart. So I think most customers are realizing we're in for a new wave of complexity, a new wave of volume. We've got to get this right because the value line is moving up the stack. It's no longer in the infrastructure layer. This should be dial tone. The value is the developer services. It's being able to quickly spin up AI languages, which is why we're interested in private AI foundation to make that turnkey super fast. So all about automating the stack and breaking down the silos, that's where I'd say is the heaviest lift in driving and delivering consumption.

Dave Vellante

>> Well, it's a strong message, I mean, especially we heard Hock Tan tell us a story about how Broadcom spends 1% of its revenues on It Many companies spend 5, 10, 12, 15% on it now. I'm curious as to how that resonates with... I could see a customer saying, "Yeah, but Broadcom's this engineering firm It's got very low, SG&A. It's got 60% operating margin and it's a very efficient company. We're in a different business. But I think the messages you could still get to that dial tone level as is in the two B can be better than it is now.

Michael Gannon

>> I'll give you an onboarding experience coming from VMware and then being a part of Broadcom. Day one. We moved the entire Microsoft productivity tools, the G Suite, day one. We do not have an IT help desk I can call in and say, "Hey, I'm having some trouble. It's a chat bot." So don't ever underestimate the ability for a human to evolve into a new operating model.

Dave Vellante

>> People in process.

Michael Gannon

>> People in process. Again, the simplicity of our business model, the simplicity at which we run IT, there's a lot to be learned by that. We migrated five ERP systems in six months time to Oracle. Six months time. We had five ERP systems lingering around VMware we never integrated. It's immediate. It's extreme execution. It's a focus on simplifying the environment so that we don't have bloated costs. There's so much to be learned on just how we run IT, and that's why Alan Davidson's a wonderful asset of a person to speak to and says, "Here's how we run it. We simplify things, we make it easy, and we don't need huge IT costs."

Dave Vellante

>> I mean, it's ironic. Everybody talks about lock-in, the fear of lock-in. There's so much competition in the technology industry. I think he would say that a homogeneous environment is going to get you a better TCO than having a mixed environment.

Michael Gannon

>> You're either locked into a technology or you're locked into your people one way or the other, and people are probably the most unpredictable asset in the shop, right? A computer's going to be always predictable for the most part.

Dave Vellante

>> And potentially the least resilient. But what do you want to be able to say? Last question. We're out of time here. What do you want to be able to say, Mike, a year from now that you can't say today?

Michael Gannon

>> A year from now, I would say we proudly have delivered on the vision. I mean, we are executing on this today. We're nine months into the acquisition. We've already stood up 158 customers around Cloud Foundation. I'm going to be doubling those numbers next year in my strategic customer base, which is the largest Fortune 500s in North America. So I'm going to proudly look back on all the value we delivered similar to if you do a backward lookup of all value realization we delivered to this market with vSphere, hundreds of millions of dollars were saved. I want to look back and say trillions of dollars have been saved by driving virtualization deeper into the stack and giving customers exactly what we told them we were going to deliver on.

Dave Vellante

>> That's very clear. Good luck with that. And Stephen, any research that you want to share with the audience that you're working on? How to get in touch with you?

Stephen Elliot

>> Yeah. IDC.com. We've got research on every theme around the world relative to technology.

Dave Vellante

>> Anything personally you're working on that is exciting?

Stephen Elliot

>> We're always working on things relative to repatriation relative to public cloud. I mean, management operations, C-suite, silo, adjustments, leadership, so yeah,

Dave Vellante

>> Repatriate or not?

Stephen Elliot

>> Yeah, I mean... Yeah.

Dave Vellante

>> Okay.

Stephen Elliot

>> Look at it.

Dave Vellante

>> All right. Go to Stephen, check out all the data, IDC.com. Mike, Stephen, thank you very much.

Michael Gannon

>> Thank you guys. It was a pleasure.

Dave Vellante

>> . Appreciate it. All right. Keep it right there, Rob Strechay and I will be back. We're wrapping up day three of VMware Explore 2024. You're watching theCUBE. We'll be right back.