Hi. Hello, everyone, and welcome to Vertiv's 2026 Investor Conference. We have a full room here today in Greenville, South Carolina, and many more on the webcast. Welcome, everyone. First, let's take care of a quick housekeeping item. I would like to point out that during the course of this event, we will make forward-looking statements regarding future events, including the future financial and operating performance of Vertiv. These forward-looking statements are subject to material risks and uncertainties that could cause actual results to differ materially from those in the forward-looking statements. We refer you to the cautionary language included in today's presentation, and you can learn more about these risks in our annual and quarterly reports and other filings made with the SEC. Any forward-looking statements that we make today are based on assumptions that we believe to be reasonable as of this date.
We undertake no obligation to update these statements as a result of new information or future events. During our conference, we will also present both GAAP and non-GAAP financial measures. Our GAAP results and GAAP to non-GAAP reconciliations can be found in the investor presentation found on our website at investors.vertiv.com. All right. Let's take a quick look at the agenda that we have. Starting us off, Vertiv CEO, Giordano Albertazzi, will be talking about how Vertiv is shaping the industry. Following his presentation, we'll have a quick 10-minute break. We'll come back, and Vertiv CFO, Craig Chamberlin, will take the stage. We'll do Q&A after that. We have a full afternoon of programming. Then we get to continue this day 2. It's a 2-day event.
Tomorrow, Scott Armul, our Chief Product and Technology Officer, will take the stage with a presentation focused on how we are shaping the industry with our technology and innovation. Super excited about that. We'll have Q&A also following Scott's presentation tomorrow. All right. Joining the stage momentarily will be Vertiv CEO, Giordano Albertazzi, after a brief video. Again, welcome to Vertiv's 2026 Investor Conference.
A profound shift is reshaping the digital world. AI has transformed data, driving unprecedented speed, density, and scale. Infrastructure is rapidly evolving to drive a new generation of AI factories across the world. The next era of compute demands a new level of system performance. Vertiv is driving the next step. Converged infrastructure where power, cooling, control, and services operate as one, delivered as integrated repeatable systems built on decades of experience, powering and cooling the world's most critical digital infrastructure. Designed as a unified system, acting as one. Power flowing intelligently across the powertrain. Thermal environments adapting to extreme density. Control synchronizing performance. Services delivering resiliency. A new generation of intelligence built for multiple compute generations ahead. Vertiv, orchestrating converged infrastructure for the AI era.
Welcome. Welcome, everyone, remote and in person. Thank you for being with us. It is great being together again after what is one and a half year exactly from our last investor event in Atlanta. It's been a very eventful year and a half. A lot of things happening. The industry accelerating, the technology accelerating. A lot of growth for Vertiv, a lot of additional strength for Vertiv. We are shaping the industry. We are the thought leader in a central industry in a big, growing market. We have the most complete portfolio and expanding that portfolio. We're scaling rapidly and in a disciplined manner to make sure that we enable the industry. In these 18 months, we have delivered, indeed over-delivered on our 2024 trajectory, the trajectory that we shared with you in Atlanta. We have achieved investment grade.
We have certainly expanded our valuation, and we are now in S&P 500, and we are accelerating further. We are leveraging our portfolio. You will hear me talk about the product, the portfolio, the system, the converged infrastructure, the services layer. We are expanding our addressed market, and we are delivering for the AI factory, AI data center of the future, and we are scaling at unprecedented speed. We are an industry leader, and we deliver with consistency on or above our commitments, and we're poised to continue this growth. When we look at the shareholder returns we have created in the last few years, was that momentum, and that delivery is certainly represented here. Yes, some bumps in the road. There will continue to be some bumps in the road. That's the type of market we're in. Certainly, quite pleased with where we are. Not satisfied.
You know my motto, but it's absolutely central here. We look at this as our starting point. Every day, for us, is a starting point. We know we can be better as a company. We know we can continue to expand our portfolio. We know that we can continue to change the game in the industry. We will never relent. Of that, you can be sure. It's 3 and a half years, almost 4 years, since we launched our 5 strategic priorities. Those 5 strategic priorities are behind the value creation that you've seen in the previous slide. They're behind the value creation and the value that we deliver to our customer every day. It is a strategy.
It is a set of priorities that are deliberately generic because they are guiding us in the right direction, and they're absolutely relevant for our future as they have created alignment and convergence of our action in the last three and a half, four years. It is customer focus, creating value for our customer, absolutely central. Creating value for our investor, absolutely central. That's also through uncompromising financial strength. It's about centrality of innovation. It's not just having a lot of new products with new technologies. How everything becomes a system and new ways of interpreting and driving the industry. It is continuing to work in operational excellence. It is delivering on what we say and continue to lift the bar for ourselves, and it's an environment, a culture of high performance that drives us.
When I look at those strategic priorities, I say, "Well, we've done a lot." More importantly, there is so much more opportunity out there. This will continue to drive us as it has driven us in the last 3, 4 years. Let's take a quick look at Vertiv, a snapshot. A snapshot of our mix in 2025. Let's start a little bit looking at this year. We are guiding $13.75 billion revenue. That guide, certainly a lot of growth, a total 34% growth. You see this growth represented in the number of employees in field services, importantly, field service employees in total employees. This is predominantly manufacturing service innovation that is being unfolded, fueling current and future growth. If we look at our portfolio, we continue to see power and thermal being the biggest parts of our business. Power management, thermal management.
We see service as a very important part of our mix, and you will hear us. We are very bullish in the direction in which service will continue to grow, and c ertainly, IT systems. It's important to talk about infrastructure solution. What you see here represented with the infrastructure solution, SmartRun, Vertiv OneCore, is the added value, the value addition around prefabrication, around manufacturing, everything that is pulled through from a service, power management, thermal management, and IT systems is represented under their various business. Don't be misled. Infrastructure solution is much bigger than you see here. We're continuing to expand our leadership. The momentum is strong, and that's why we are pretty confident in our ability to continue to deliver on our trajectory that we shared with you in 2024, but even more than that. We have demonstrated ability to deliver.
We have demonstrated ability to exceed our financial frameworks, and that's what we will continue to do as we look into the next five years. It is a multi-year executional strength. I want to take a few moments to look back also in terms of our growth in top-line margin, cash, and adjusted EPS. Top line, 24% CAGR in the last five years and accelerating as we have seen in 2025 and 2026. That, of course, has driven margin expansion. It doesn't just happen, by the way. It doesn't happen, just happen. It's a lot of hard work in terms of value creation, in terms of margin expansion. It is certainly also operating leverage. Certainly, it's nice to look at 32% margin expansion CAGR. That, of course, translates into EPS acceleration 12x the EPS that we had in 2022. Certainly a lot of strength.
Generating $2.5 billion, more or less, cash delta between the extremes, and it's something that fuels, of course, the strength of our balance sheet, and certainly that enables us to make the right choices when it comes to capital allocation. All this, of course, is a result of an absolutely focused team across Vertiv, but certainly within the leadership team. Let me go through the leadership team at Vertiv. Some changes, certainly, since we were together in Atlanta in November of 2024. Indeed, six roles that are different. Some are people from within Vertiv. Paul, who's here today with us, now leads our EMEA business. Scott, that most of you know already, is our Chief Product and Technology Officer. From the outside, Wei, now leading China.
Of course, Craig, who you know, I'm sure, pretty much in person, everyone here, leading 6 months now on the job as our CFO. More recently, Frieda, Mike, sometimes mid last year, 6 new names or old names in new roles. I'm very pleased with the team. I'm very pleased with the strength of the team. Everyone in the team combines very strong strategic vision with an obsession of execution, and an ability to really do 2 things, be ambidextrous in terms of the strategy and execution. It's a team that is driving our strategy forward.
As we think about strategy, as we think about a strategy based on the vision of the team, the vision of the board, and the principles that I was sharing with you earlier, we continue to go back to what we believe is a very proven framework for value creation, proven but very future-ready as we go through our strategy, as we look at the market, as we look at the technology and how it evolves. Center around continue to strengthen our leadership in a market that is favorable. Continue and strengthen and accelerating our role as an innovator in the industry. continues to make sure that we have the most complete portfolio. It's not static. You know it very well how many changes, how many evolutions this portfolio and indeed the technologies that we serve, the data center technology, the range of data center technology have expanded.
That's good news in many respects because that makes our portfolio net broader, not just different, much broader. That, let's say, bigger spectrum of technology makes the market more complex, and complexity is our friend. Extremely important in this very dynamic market is the relationship with the ecosystem, and the ecosystem are key other players, the leaders of the world, partners, but also our customers. The strength of the relationship and the long-standing relationship with our customers is absolutely key and something that we nurture every day. Then there is the executional part, having a clear and continue to work and to continue our trajectory and a journey on a roadmap to operational excellence. It is continue to expand margin through operational leverage, but a lot of efficiency, as you will hear from Craig, and certainly a lot of commercial excellence.
That generating the cash flow that then becomes a strength of the balance sheet, that then becomes an ability to own our destiny in our own hands. This is a very consistent framework. You will kind of see this framework unfold as we go through my presentation. I want to start with going through again, because you've seen it before, and the beauty is that our competitive advantages is something that we continue to reinforce. That value creation framework is exactly reinforcing this set of competitive advantage. It's a uniquely deep application expertise. We know the data center space like no other, and we've done that for a long while, but never believing that what is true today will be true tomorrow. We not only adapt, but we shape the future of the industry. Customer collaboration, very deep.
We are shaping the AI data center of the future. That starts with very strong relationships. Most complete portfolio. We will always drive that angle. Relentless innovation. We will always strive for the best technology at point product level. Think about densification. Think about also at a larger scale, how we're rethinking the way the entire infrastructure can be deployed at speed, at scale, entire infrastructure as one product versus Vertiv OneCore. Proven reliability and quality. Many aspects to that. An important one is the fact that we have a very large install base. The amount of intelligence we get from that install base enables and arms us for future technology development. All things that reinforce, that accelerate the flywheel of our competitive advantages. We're truly global and operate as a global company, and we provide consistent experience across the world, and we are able to scale. We're proven to be able to scale.
Certainly last but not least is the strength of our services. As the industry becomes more complex, significantly more complex, not just in the technology but in the way, the speed, the scale, the demands are evolving, services are central across the entire spectrum of our service portfolio. You will see that spans from the initial phases of almost design of an infrastructure, all the way to the entire life cycle of that infrastructure. We actually enable our customers and enable the industry navigate the current and the future challenges. With that strength in mind, with that strategic clarity, with the strength of a team, with our proven track record, we are then reviewing our five-year projections, and we are lifting our trajectory relative to what you saw 18 months ago.
We now think in terms of a top-line growth in the 5 years on the 20%-22% CAGR. It is above market growth. We will go in the details of the market later on. Continued data center leadership, very strong in the key customer segment, and we continue to see growth across all aspects of the market. We believe our margin expansion has a rightful ambition on the 27%, and Craig will elaborate further on this. Again, it's operating leverage, it is efficiency, it's service acceleration, and it's certainly strong commercial execution and continued enduring strong cash generation. Disciplined execution, improved profitability, driving stronger cash flow. As we continue to expand, what are the forces behind this expansion? I want to go back to the theme of shaping the industry, and how are we shaping the industry and creating enduring value.
I like this picture with these concentric circles because really it represents how we look at the market as our offering. The market is favorable, but the favorability is not just about the net growth of the market. It's the fact that the industry is changing rapidly, and as it changes, it creates, I like to think, a fertile ground for our competitive advantages to really work extremely well and to create value. We create value at product level, and that's the second circle. At product level, product leadership. We will always strive for being the leader in any specific product line in which we operate. That means that we want to be at the top from a technology standpoint, and we want to make sure that also from a market in that specific part of the market, we continue to lead. You will hear me talk about system.
You will hear me talk about converged infrastructure. Do not make the mistake to believe that we will not continue to win on a product-by-product basis. We will continue to lead and win at a product-by-product basis. There is another layer that is the system excellence. There is so much more value in the seamlessness of a powertrain, a thermal chain of an IT solution when it's really defined, designed, thought through together. You know that that's something that we do a lot with our customers. We make sure that we sit with our customers early on in their process and make sure that we talk the entire system and the advantages. We want to make sure that our customers can benefit from our knowledge of the space, of the domain, to optimize their infrastructure.
Sometimes they will buy the entire system, sometimes they will buy one piece at a time, and sometimes they will be multi-vendor. It's okay. We want to make sure that they understand what we understand being excellence from an infrastructure standpoint, and sometimes it will be a converged solution. We move to the next level, that is the entire infrastructure. When it is the entire infrastructure, things change further. You really think about what are the various forces that we have to take care of in the market, and I will go there. For now, really think in terms of the whole infrastructure as one coherent, rapidly deployed system product. The entire infrastructure is one product. That's what OneCore is about. That's what SmartRun is about. There is another layer, another ring around these concentric circles.
That is a ring that embraces them all, almost, I sound like Tolkien here. It is one ring that embraces them all that is service. You can understand how complex, you can understand the speed, you can understand the technology. That technology speed and complexity need superior service capability. Let's start with the first of the various circles, and let's start with the market. This market is certainly giving us very interesting and, dare I say, persistent signals. Certainly, one is compute growth. I'm thinking of we're talking about the next five year, 2025, 2030 range. This signal continues to indicate strong compute growth. Now we're talking about predominantly, in terms of the fastest-growing element here, we believe we see is around inference, almost at 40%. That's very good.
We see inference just like training and, let's say, traditional compute as areas where we can help create a lot of value for our customers and for our investors. We like it a lot. Data center CapEx is growing. When we think about the total power capacity added in the five-year, we think it's never exact, but it's a projection. We believe that what was about 100 GW a year and a half ago, it's looking more like 140 GW. Remember, we're always quite balanced in these analyses. We are not negating our personality, let's say, measured and balanced as we do this here. Think about increments that go from approximately 20 GW in the first year to somewhere like 35 GW directional. Again, is the appetite bigger than that in terms of growth? Yes, the appetite can be certainly bigger than that.
There are external forces. Some of the external forces we've been dealing with for quite some time. Those pacing items really are around power availability, permitting, and now even more so skilled labor, field capacity and field complexity in field, all things that are pacing the industry. They're not stopping the industry. We've seen that being true in the last almost four years now. If we look at the market, I'll start talking about our legacy serving market. What does that mean? That's pretty much our 2025 portfolio. It's a $62 billion total market, growing 16%-18%. The data center represent about $50 billion, growing 18%-20%, more or less, with the cloud and colocation being the fastest-growing part of that, about $30 billion growing at 23%-25%.
Just to give you a sense, our growth in cloud and colocation last year was about around 45%. Enterprise and distributed IT growing a little bit faster than we saw when we were in Atlanta. Don't be misled. This is really talking about enterprise and distributed IT on-prem. There is much more AI adoption in enterprise that happens on a cloud, certainly, but also on a colocation basis. We are expanding our market, and we are increasing our TAM by about $13 billion. Think about a $75 billion growing 16%-18%. Those are all the technologies that we are adding and parts of the market that we are increasingly serving, being those organic or inorganic. Think about fluid management, PurgeRite. Think about IT solution expansion, again, acquisition.
Think about all around the converted infrastructure that is changing the shape of the market. Think about the evolution of the power train that is driven, among other things, by the DC power in data centers. Think about the chilled water and switchgear technology that is being expanded globally and regionally. Market is expanding, but there are 5 forces that are intensifying, that are driving this change. Density, speed at which you make capacity, compute capacity available, scale constantly increasing, complexity. Navigating all the various technology, all the various designs, and certainly more and more complex load designs. Add to that, a lot of power generation happening, dare I say, within the perimeter of the data center. It's easy to understand that the traditional old way of designing data center, of delivering data center may not be optimized.
There are a lot of seams connecting all the parts that are sub-optimized, if you think about a data center design and build one technology at a time, one piece at a time. There is a lot of value, cost, profit opportunity, time trapped in those seams. That's why more and more we think that the industry can change and is evolving from the point of view of how data center are developed, deployed. What are the forces that are driving the data center of the future, the AI factory? We like to think in terms of, and not just us, the industry increasingly so, in maximizing tokens per second per megawatt. The real scarce resource is power. How do you maximize power, the power that you have available?
How you make sure that all the electrons that you contract, or all the electrons that you generate, if you are self-generating, are converted into compute, and then from compute into tokens. We like to talk in terms of overall throughput, in terms of second, tokens per watt efficiency, tokens per dollar, how the capital is really returning of an AI factory or data center, and what is the time to first token. That's why you see a number of technologies that are really being developed, not just by Vertiv, but certainly with a big role of Vertiv to address those elements. That's why you see higher voltage power architectures, example. That's why you see more hybrid liquid air, hybrid heat rejection structures. That's why a lot of modularization and convergence. Let's go back to Atlanta real quick, and think about what happened since then from an innovation standpoint.
Innovation is central. We want to stay ahead of the industry. We will go through 120, exact number, 120 innovations, be it products, be it controls, be it services, you name them, that we have taken to market in this 18 months. It is in power management. It is power converters that can, best in industry, we believe, handle the spiky loads of AI, dynamic loads of AI. It is about battery energy storage systems. Is it thermal continuing to expand the portfolio, liquid cooling in chillers, in air, new products, new launches. It is changing the game infrastructure solution. If you think about Vertiv SmartRun, a tremendous success, and changed the game in the industry with Vertiv OneCore. IT system expanding, services, super important. Our Waylay acquisitions really powered our ability to bring digital and AI into service execution with predictive maintenance.
The acquisition of PurgeRight opens a new part of the business, I would say to us, a new market. Think about an innovation, some of you might have had words with Ron already. Think about the Near Zero, a way to really do the commissioning of a data center, saving an enormous amount of water. Very important. Software with layers that are system level, Vertiv Unify. I could go on for hours as you, someone will go on for at least a good hour tomorrow. With that, innovation absolutely central to all we do. Let me go back to my concentric circles. Again, I want to show this as layers of value. If you think or anyone thinks that it's enough having the product range, the portfolio, and being a leader in each product, and we will be, we'll continue.
We're not being distracted there. That's not enough. It's great. It's absolutely necessary and great, but it's not enough. It's about starting to add a layer of a system, a layer of converged infrastructure, and the service ring that rules them all. With that, let's go straight into converged infrastructure. Let's hear directly from Asher Genoot, the CEO of Hut 8, that talks about the relationship, the ability to scale, and the innovation and the importance of OneCore.
One of the key reasons that Hut 8 chose Vertiv is because of the infrastructure and the platform and scale of their operations today. As we chose a partner, it wasn't about a partner that could just scale with us at our Riverbend campus, but also a partner that could scale with us reliably and be on schedule across many more projects. Hut 8 believes that this is one of the most exciting times in the data center market. In the recent years, we've gone to direct liquid cooling technologies.
As we look at Hut 8's partnership with Vertiv and being able to redesign and rethink how not only do we deliver the engineered solution for the data center, but what component and what parts of that are done in the factory versus done in the data center itself, we think unlocks a ton of opportunity for advanced acceleration and opportunity for innovation within the delivery and the solution of the actual infrastructure stack for the customer. When we approached our Riverbend data center, the goal was how do we deliver a large-scale infrastructure stack in a manner that was able to de-risk project execution and take a lot of the QA and QC process and systemize it.
By being able to have a OneCore modular design, that allows us to be able to take a lot of the risk off of the site into a controlled environment, into the factory, and to be able to deliver skids on site and allow those to be integrated on the campus rather than the thousands and tens of thousands of points of connection that otherwise would have happened on site. Hut 8 believes that when choosing a partner within the supply chain, it is not just about who can deliver infrastructure to you up front, but who has the ability to actually manage and maintenance that infrastructure and to stand behind the reliability of that equipment. With Vertiv, our expectation is for them to stand behind the quality of that equipment and the servicing of that equipment over the duration of our delivery to the tenants.
The key in growing and scaling is the ability to rethink how you're building, to rethink the problem that you're solving with the quick acceleration and advancement of the technologies and the compute that supports those technologies. Hut 8 believes that as it continues to grow, it needs to find partners that are willing to innovate alongside them, like Vertiv, in the whole stack, from supply chain to delivery execution to end delivery to the customer.
Thank you. Thank you, Asher. I think it is about innovation, about thinking about the entire supply chain differently. It's changing the paradigm in the industry. Partnership and ability to scale and having a design that is fully optimized from day one, avoiding all those points of connection that are tricky, especially when they happen on site, is absolutely central. If we think about OneCore, if we think about what you heard here, is what we call converged infrastructure. Is when the entire infrastructure is optimized. Now, let's not forget about the fact that it's multilayered. The other slide, product, system, converged infrastructure. That's fundamental. We play all the levels. We will never defocus from the day-to-day, product-by-product, leadership success and technology. OneCore delivers elements that are absolutely fundamental, especially through the lens of the forces that we described earlier in the presentation.
It is up to 50% faster deployment time. It is really when you move field to factory, you can also increase the quality, the reliability, the speed also in that respect. It's not only how fast you are in deploying, but think about how faster the commissioning of a very complex system is if it's done in de facto in a factory. Think about the reduction of footprint and an optimization of tokens per dollar of an infrastructure that is fully designed and optimized as a whole single product. When that happens, and it's not a big part of the market yet, we believe, but when that happens, then the ladder of value when it comes to building a data center shifts, as we shift from field to factory.
Clearly, at that stage, a Vertiv converged infrastructure delivers value to our even day one value, let alone during the life cycle, value for our customers and enables us to capture a bigger share of the overall spend for data center deployment. Again, you heard it also from Asher Genoot, it's not just what happens when you build the data center, but it's also what happens during the life cycle of the data center. What happens before, during the data center building and commissioning, and the life cycle is really the domain of Services for us. Services is definitely one of Vertiv's superpowers. With that, I'll leave it to this video to elaborate further.
AI workloads are redefining critical infrastructure, changing how systems are designed, deployed, and operated for performance at scale. Vertiv Services is a lifecycle model built for recurring value, with expertise that turns complexity into confidence and delivers consistent performance across the full lifecycle. Consulting starts with engineering insights that shape outcomes before deployment. Design optimization is informed by real system requirements, driving efficiency, scalability, and long-term value. At the implementation stage, deployment excellence happens at global scale, with commissioning built on OEM knowledge and proven standards, so customers get consistent execution, repeatable quality, and a faster time to value. With always-on service support across the global installed base, lifecycle services sustain performance over time, even across complex mixed-vendor environments. Through continuous performance optimization and operational insights that improve efficiency over time, intelligent tuning can deliver up to 15% energy savings.
With Vertiv Next Predict, predictive maintenance is informed by real operating data, providing actionable intelligence across the infrastructure portfolio. Vertiv Academy build expert teams for consistent global execution with training aligned to real-world conditions. Vertiv is uniquely qualified to support AI infrastructure at scale with approximately 5,300 service engineers, more than 31 training academies across 6 continents, and approximately 320 service centers. Vertiv Services, powering AI performance across every phase of the customer journey.
Every phase of the customer journey. We are scaling our services very rapidly and convincingly. You can't scale if you do not have the critical mass, if you don't have the preexisting strength of presence on the territory, and if you do not have a proven network and well-working network of academies, training centers, as we are. That is the entire lifecycle from the initial phases in consulting, in which we make our know-how available to a customer in helping them to optimize what they are designing to the implementation phases. Commissioning. It is a start-up. Think about how complex a modern data center is. Think about how many pieces need to come together. It's like tuning all the instruments of an orchestra. That can be very complex and very time-consuming if not done very well and very professionally.
You have to be first time good because any delay means a delay on your returns on the capital that you're putting in the data center. The lifecycle of the infrastructure of the data center begins. That's where our customers really see the productivity of their asset. That's where it is the maintenance. It is as the data centers are becoming larger and larger is its on-site presence. It is where, dare I say, optimization and digital and life cycles start to blur because the technology creates Next Predict. Technology gives us preventive. Our network of operating centers helps remotely monitoring. Also during the lifecycle of a data center, a lot of things happen. There is a lot of readjusting, retuning constantly that instrument.
If you think about the cycles of IT that go through a data center or a data center goes through, that requires every time to retune. It's the odd server swap in a liquid cooling, high density that creates possible imbalances that needs to be looked after. All reasons. That's the complexity I was talking about, and that's the complexity that makes us and our services are so central. Look at it, an example. Let's take a look at the thermal chain end to end. Of course, we consult. We consult on what the piping should look like. We have experience, ton of experience there. We help what the flows should be, the liquid flows should be, then comes the commissioning.
When you commission a thermal chain, and the same is true for the powertrain, for all the parts, but I'm using the thermal chain as an example. You commission, you have to make sure that all the CDUs, all the air units, all the chillers or heat rejection units outside are tuned exactly to work with the rest of the system. Then you have to work on all the fluid management and balancing. Extreme cleanliness. Then the life cycle begins. The data centers are the size as such that you have to have a permanent crew just to make sure that all the normal maintenance takes place on a regular basis. That's why, again, if you have your entire thermal chain made of Vertiv equipment, you can deliver certainly some leverage there. Then you constantly optimize.
Vertiv traditional services is absolutely fluid management at its best throughout the entire life cycle. We create value for the customer. The install base is growing. The install base is becoming more complex. That creates a lot of recurrence, because that is install base that would capture. The more complex is the install base, the higher the capture rates, and that certainly is a circle that we like a lot. We know that we create a lot of value for the customer, and we create a lot of value for our investors. The install base creation is accelerating, and we're getting ready for that, and we keep expanding our services capabilities. Put everything that at a very high level I shared with you in terms of where the technology is going, and that takes us to our current view of the TAM per megawatt.
It's just a couple of caveats here. One is, this is a snapshot in time now. The other is, this is influenced by the mix in the market. It's clear that when we talk about converged solution, if this is the spectrum, we are way north of the spectrum. This is the weighted market that we represent here. One thing is sure, that we believe that going forward, the dynamics in the market, the dynamics in the technologies will drive further TAM expansion. Now, I was talking about our value creation framework. A lot of emphasis on the relationship in the industry. Clearly, relationship like the one with Nvidia, absolutely central. Absolutely central, and I think we help a lot the industry with that relationship. With Dell, with Caterpillar, with Generate. It's not an industry where you can operate solo.
You have to be a very structural part of the ecosystem. Customer collaboration and full-spectrum expertise. When we talk about customer collaboration, I'd like you all to hear directly from Chris Crosby, the CEO of Compass Datacenters, what the relationship and what the mutual strength of the relationship looks like.
One of the great things about our partnership with Vertiv is that they're right there beside us. The type of scale that Vertiv can bring to the table, coupled with the speed and the quality, that's what we're looking for, and that's what Vertiv brings to us as a partner. Compass saw our customer needs evolve quickly, where the white space fit-out became a component of our responsibility as opposed to theirs. We wanted to bring that same manufacturing style to the data center floor that we had done to all the other components of the building. We went to Vertiv to help us to do that. That's how they came up with a SmartRun to really serve the future as well as today in the white space. One of the best stories that we have with Compass and Vertiv relationship is Vertiv CoolPhase Flex.
Our engineers saw that we needed to get to the future quickly for AI, a flexible future that could do air and water in the same data center. We worked closely with the Vertiv engineers to come up with a solution that is really groundbreaking in terms of its flexibility as well as its performance in an AI data center. When you've worked with water before, you know that the service layer is the most important layer. The care and feeding of these water loop systems is of utmost importance, especially when you look at the sizing and scale of the investment that goes into the data center itself. We've got to protect that more than anything else. It's fantastic that Vertiv has developed this service layer, and we plan to continue to exploit it to the benefit of ourselves as well as our customers.
PurgeRight was a company that Compass had selected to help us with flush and fill activity, which is so important within a liquid-cooled environment. We were very excited to hear about the acquisition of PurgeRight by Vertiv. That gave us the scale and the comfort that we felt necessary in order to go ahead and roll out with a national program with PurgeRight. We're all in on NextPredict. We couldn't think of purchasing a piece of equipment without the maintenance and the predictive analytics to go with that. It's a bumper-to-bumper type of mentality that we take here at Compass, and we want the best people in the world maintaining that equipment. In this case, that's Vertiv. Vertiv's embedded in everything that we do, whether that be our design, our specifications.
They even have a cube in our offices where their employees can sit to be embedded with our supply chain relationship managers. This is a core relationship for us, and we are so excited about what the possibility brings when we look at everything from the tip to the tail. How can we get better, what can we learn, and what can be the future?
Thank you to Chris and the entire Compass team for their partnership. A really profound partnership here. A partnership, of course, that is grounded on technology, on services, on a different mutually continuous change of the way we look at things, how we can make the industry better, how we can change the paradigm. That is so core for how we look at the business in general. It's also based on the ability to scale. Scaling, we have. We have been scaling, and we have done so in a disciplined but decisive manner, and we will continue to scale. It is about executing on the CapEx investments. It's about implementing and leverage our operating system. There is a lot of productivity that has come to fruition. There is a lot more productivity that is there that we can extract from the system.
We will continue to expand with safety, on-time delivery, being focused on the new product development, and introduction. We're talking about 120 innovations, and certainly continue to evolve our AI utilization in manufacturing, in general, advanced manufacturing. Again, let us go back to Atlanta, and let's see what has happened in the last 18 months. All the capacity that we have added. Again, the axes are more footprint, more technology, and certainly a lot more efficiency productivity from what we have. These are the axes that characterize our trajectory here. It's a lot. It's a lot of also new capacity and expansion. In South Carolina, we have 3 infrastructure solution factory very near one to the other. We'll visit one of them. It is in Ohio. It is in Pennsylvania. It is in Europe. It is in India. It is in Asia. It is in Mexico.
The entire range of capacity is being expanded. This is only in the last 18 months. A lot of focus there and a lot of discipline. By the way, we're not saying that we are doing things to perfection. We have so much we can improve across the board. This is true for everything. Certainly, there is a lot of leeway in our operational excellence trajectory. Let's go back to the 2030 projections. Certainly margin expansion, top-line growth, margin expansion, strong free cash flow, strength in our balance sheet, cash, and an ability to continue to focus also organically and inorganically, also on M&A as a source of strength and of growth. You saw us in the last five years make nine acquisitions.
Certainly an acceleration in the last less than a year with six different shapes and forms across all our businesses: thermal, power, infrastructure solutions, IT, service. We like service a lot. You will hear it. I'm sure you will hear it from Craig. We like service a lot. Again, sometimes it's adding new pieces of business, like net new capacity go to market, customer access, technology, like we've done with E&I, like we're doing with PurgeRight. Like we have done in many respects with Great Lakes. Sometimes it's entering a technology that at the very early stage, but it's the right time, the right technology, then we scale. We scale like we've done with CoolTera and everything liquid cooling, as an example. That's a modus operandi that we like a lot, but not the only one.
When we think in terms of our playbook, we believe that our playbook is strong and getting stronger, both in terms of acquisition, but also in terms of integration. Our pipeline is robust, and it's quite vibrant. The criteria behind our M&A action are pretty much what we've been sharing with you for quite some time. It's either technology differentiation. It's either market access. Sometimes it can be capacity. Sometimes it can be adjacencies, the whole lot. What is important is that we create or have a potential to create above-market growth rate, growth margin accretion, and certainly reinforce our value creation equation that then you have seen in the 5-year projection. Feeling stronger and stronger in this respect.
All what I have shared with you is really giving us the confidence in delivering on our long-term financial projections with above-market organic growth between 20% and 22%, a margin ambition of 27%+ . Strong cash flow, strong ability to deploy capital, certainly staying within our debt leverage framework and continue to deliver for our customers and for our investors. With that, thank you for your attention. 10-minute break, then we're back with Craig. Thank you very much.
[Break]
All right. Welcome back, everybody. Thank you. Joining us next on stage is Vertiv CFO, Craig Chamberlin.
Thanks. Welcome back, and thanks for the applause, and thanks for being here. We'll take a little bit of time to walk through the financial slides, and then I'll call up Gio and Lynn. We'll do some questions. Hopefully I can get through this so we can get back to the money part you guys want to do is ask us some questions. We'll get through that. When Gio's been talking about it and we've been thinking about it, we have five key areas for value creation. Gio hit on a lot of them, but I just wanted to reiterate them. The strong execution drives our financial performance, and we have a track record around this. You saw it from 2022 up through 2025, and then we're going to continue to be able to do that through 2026 through 2030.
That really drives the meaningful upside that we see in the future here. Above market organic growth, you see that whenever we're talking about we see the market growing at 16%-18%. We believe with our technology and our market, we'll be able to get that to 20%-22%. Adjusted operating margin expansion. I'll go into this in detail, but really what that is going from the 23.3% to the 27%. That margin expansion we see along the time horizon. We really believe we can drive that. Now, again, I'll double-click that as we go through the individual slides. The adjusted free cash flow conversion, Gio talked about it, and I'll continue to hit on it. The profit engine that drives the cash, but not just the profit engine that drives the cash.
We like to also really focus on the operating mechanisms that drive the cash, whether that be operating cash flow through inventory and through payables, all the way down through the customer cash cycle, driving milestone billing payments so we stay ahead of cash. Also on the side, just making sure we get the right collections status so we can continue the cash flow. That all encompasses itself with this flexible capital deployment.
We really want to be able to invest in the business going forward, whether it's in technology in ScottSpace, whether it's in M&A to help us continue to grow, whether it's in all the capacity that you saw populate up through here so we can continue to pick up the demand that our customers are giving us into our services business that you heard about with Chris and PurgeRight and all the great things we're doing there. All of that flexible capital deployment gives us the money to fund the business and the growth going forward. I get this question a lot, so I thought I would hit on it here. In your six months, what are you going to focus on? Six months going forward, where do you really think that you want to put your fingerprints on the business?
I kind of laid it out in three different areas. One, strength and execution. We have the ability to lever as we go and grow on our fixed cost. One, we want to make sure that that delivers margin expansion as we deliver out, and that we continue to be able to be on time for our customers. Also within that is productivity gains. How do we make sure our factories are flowing so we get that variable cost leverage? How do we make sure that we're getting the best out of our material cost leverage? How are we ensuring that our shops are lean and they're automated? We'll go into a little bit more on that too as we go through the pitch. Growth services. I get this question a lot and we talk about it a lot. You heard Chris Crosby up here.
Services is a main portion for what our customers look to us to be able to do, and we have a massive install base. We want to be able to mine that install base, and that will help future growth for us, and it's a big future growth engine for us. Recurring revenue. A lot of people think of that as just parts and break/fix. As you heard when Chris Crosby was talking about it, Vertiv Next Predict, optimization, all the other underlying services businesses, the consulting, all that is a flywheel. It's not just parts, it's not just fix. It's also optimization. It's also consulting. All that creates a flywheel back to then you being the customer of choice or the supplier of choice whenever they're trying to solve new problems on the OE side. We love that flywheel. Generating cash.
Cash, cash really is the lifeblood of our business. Improving profitability, we talked about that. The first two will do that. Operating working capital. How do we stay focused whenever we're delivering on making sure we have a very, very good operating cash cycle? Meaning lean on inventory, still being able to deliver for our customers, having really good partnerships with our suppliers so that we have that flywheel going and we can generate cash out of that operations. Commercial, staying ahead of the cycle. We like our milestone billing payments. We like to be able to collect cash on time and ensuring that we stay ahead of the cash so that we have ample capital to be able to deploy.
Gio hit a little bit on this. The growth rates that we expect to see, 20%-22% long-term organic growth is really fueled by our market expansions and technology leadership. We expect to stay price cost positive through the cycle. Of course, there's going to be inflation. We expect to be able to price within that and get a little bit of margin expansion on that. We'll show you what that looks like in a walk here in one second. The underlying portion of this also is we do expect a lot of pickup out of services. We have services in here at 20%+ throughout the period. It is, it's one of our superpowers. You heard Chris talk about it. You heard Gio talk about it.
It's something I bring up all the time whenever we have conversations either with analysts or on our earnings. This is the area where we have to continue to be strong when we go into the future. Talked about our 27% adjusted operating margin leverage, I broke it down into three different points here. Operating leverage, We talked a little bit about that already. We maintain our strong leverage as we're able to produce and deliver for our customers. We do want to continue to invest. That does take really good operations. If you're going to invest in technology and services and grow your underlying investment into the business through capacity, you've got to have productivity and you got to have flow of your business to make sure you can fund that and get the margin expansion on the leverage.
You'll see two points through that through the period that we expect to be able to get on volume while also doing those investments. Productivity. You've heard a lot about VOS. I'll break it down and make it real for you here in another page. We'll talk about manufacturing productivity. We'll also talk about material productivity. I'll give you specific examples of how those will flow through and why we're targeting where we're targeting to offset things like inflation and unexpected inefficiencies. Then commercial execution. We do expect positive price cost through the cycle. We expect to be able to price for our technology. We also expect there to be inflation. We expect there to be unknowns. We have it a little bit higher in here on a gross basis, but we've netted this down.
We want to be prudent in the way that we have the outlook here, but those are our 3 major factors when you think of the walk from the 23.3 up to the 27. I said I'd break it down a little bit more for you on the operational side, and I think of it in terms of 2 different buckets when I'm thinking of operational productivity. There's the manufacturing side, the manufacturing labor productivity, which, if you were to look out into 2030 for us, would be about a $2.5 billion bucket of cost. Through us driving lean manufacturing activities through VOS, robotics, automation through our plants as we invest, using things in best cost country, leveraging our footprint efficiency, we believe that we are targeting about upwards of 5% of productivity on that.
That's gross productivity. We're doing that to ensure that we can absorb inflation and labor, we can absorb inefficiencies in factories, we can absorb all the unknowns. As that drops through, it's a portion of that 1.5% that drops through. We specifically put a big target out there. I'm one of the major issues of driving this so that we can ensure that we do expand those margins. The other side of that coin is the material cost productivity. This is everything we buy. Again, if you look out to 2030, it's a $10 billion bogey out there of cost. What do you want to do to drive that cost out? You volume buy leverage, right? You have all these suppliers that you're buying things off of. How do you get leverage on that and get a couple points of productivity?
Make versus buy. Are there things that we should be making? There's things that we should be buying. Where do we get the best cost out of that? Should-cost analysis. What should this cost me when I buy it from a supplier and going to them and getting them to bring their costs down for us? The supplier cost reduction ideas, enforcing this with your suppliers, that if they want to be with you long term, they're going to bring to you ideas of how to make their individual products cheaper or for us to manufacture or design it cheaper. That $10 billion opportunity, we have it again in here at a gross level of 2.5% productivity over the period.
That, again, would offset the inflations that we'd feel, and it would help drop down through to that one and a half points of expansion you see when we're talking to you about productivity. Cash generation. Again, we're aiming in the period here to be between 95% and 100% conversion. A lot of that comes through on the profitability side, but the areas that we can control even further than that are our trade working capitals, right? Operations, and I talked about it, inventory turns, leveraging our flow-through and our manufacturing prowess to drive inventory at the right level. Also just ensuring that we have great partnerships with our suppliers so that we can drive the right payment terms.
Then commercial, I talked about setting up our milestones, getting the down payment, staying on the positive side of the cash curve, and collecting everything that's due to us. CapEx, we've hit on it a little bit. This framework, we've always talked about a 2%-3% investment in CapEx, ramping to 3% and 4% this year. This framework has it in at 3%-4%. That would be a little bit more closer to 4% as we ramp in these first couple of years, a little bit more down towards 3% in the out years, but it does support that 20%-22% growth that we've seen. It also helps us do things like the technology advancements that you'll see from Scott. It also helps us drive the productivity that I talked about on the other page.
How do we put in things that are lean lines? How do we put in robotics and automation? All of that funds the underlying growth here. We have it at a 3%-4% net number over the course of the period. As Gio mentioned, we are targeting 1x-2x on leverage. Again, it's a target. It's a framework. We did the refi earlier this year and structured out our debt ladder and went out to 2026 with that, and we really like where that sits today. Again, the targeted net leverage of 1x-2x, I would think of that as a framework. We're comfortably operating underneath that. We're also comfortably operating above that, if and when there would be an opportunity to go out and do the right acquisition.
We'd always want to stay back to the 1x- 2x and get back to the 1x- 2x, and that would be where we'd operate at, but it does have some flexibility in there. When you look at that, I just wanted to break it down to what does that mean from a cash availability for us. The $20 billion you see is the projected available cash that we generate across the period. If you lever this at 1.5x, you would have $28 billion. Look on the right-hand side of the usage of cash. We have dividends and share repurchases in here as an estimated $4 billion. That leaves you with a pot of $24 billion to go and utilize in terms of growing the business.
If you looked at a framework around how would you use that in an M&A space, we've spent about $3 billion-$4 billion over the last 4 years on acquisitions. A typical run rate of our bolt-ons would be somewhere between $750 million-$1 billion. That's probably a framework that you'd see us continue to think about as we did bolt-ons. The other side of this would be if there was a strategic acquisition that we wanted to go after and it had the right value for us, we have the leverage, and we have the powder to be able to go do that. Thinking of the entire framework here is we will be doing some bolt-ons. We know that that's something that's in our framework, but we also have the opportunity to do things that would be more strategic as we go forward.
I'll just hit back on the points that Gio hammered home earlier and that we would think of as our total framework. Again, the 5-year CAGR of 20%-22%, really underlying that is our strong technology, our strong market, our strong services portfolio, outpacing the market by four percentage points. Feel very strong about that. Operating margin, I went through all the different levers, growing that 3.5 basis points over the period. Adjusted free cash flow, 95%-100%, throwing off lots of cash, and then the flexible cash deployment.
We have in here about $4 billion of use, $24 billion of openness, and that gives us a lot of flexibility when we go think of bolt-on acquisitions and also the potential to do other things, whether it be transformative M&A or think about other additional returns to shareholders. That's what I had for this section. I think I'll call up Lynn and Gio.
All right. We are now going to open it up for Q&A. We have a lot of time for Q&A. We do have a full house. I imagine a lot of hands will go up quickly. A few just housekeeping points. We do have a lot of people who want to ask questions. Please try to limit yourself to one question. If we have extra time, we will circle back. The other thing, please remember Scott Armul, our Chief Product and Technology Officer, has a technology presentation tomorrow. We'll have Q&A after that as well. Technology-based questions you may want to save for tomorrow's Q&A session. With that, we're going to have a lot of hands. We're going to have two mic runners. I would ask, wait till you get the mic handed to you before you ask a question. We'll just go up and down the aisles.
Julian.
Thanks. Julian Mitchell at Barclays. Maybe just for Gio, really, the dollar per megawatt number that you gave, so it's about $3.5 million at the midpoint, and that's today, as I understand it. Maybe help us understand how does that split between, say, power and cooling, perhaps. When you're thinking about the outlook, what kind of growth should we expect in dollars per megawatt opportunity, when you're thinking about broad brush changes like 800 volt DC or liquid cooling, that type of thing?
Julian, just make sure. Question around what is the trajectory and what is the mix, if you will, in terms of TAM per megawatt. Trajectory is favorable. We're not more specific than that deliberately because a lot of things can happen. If you really think about increasing density from a cooling standpoint, that increased density I just want to make sure I don't steal Scott's thunder. That technology becomes more complex, more dense. Everything from service to the liquid cooling becomes more complex. If you think about the trajectory of how heat rejection will happen as the water temperatures will change, and I don't even dare go in there because that would be painful if I do it. It's fantastic when Scott does that. You see that even everything that is heat rejection will be more complex. You look at the power part of the portfolio.
Think about all the on-site power generation, the consequences downstream from that power to the chip. Think about the AI loads and the dynamics there, and how complex that power train is, regardless if it is an AC power, a medium voltage AC power, or a full end-to-end DC power. Think about the convergence that I was talking about. Even simpler level, kind of a more partial integration, all elements that add. We're not specific. There are a lot of dynamics. It's not just the technology, but also the mix of technologies. As the converged part, the Vertiv OneCore data center, Vertiv SmartRun data center as a product becomes a larger part of the market of the mix than the TAM per megawatt will expand.
When it comes to the mix of that TAM, I think that an accurate, a good way to take a snapshot to that is really take a look at the mix that we have today. As I said, and I had a couple of questions, in that mix the thing that is a little bit underrepresented is the infrastructure solution. When we talk about convergence, when we talk about SmartRun, when we talk about OneCore, that part is within infrastructure solutions. What we capture in that slice of the semi donut, if you will, that slice of semi donut don't include everything power, services, thermal. That goes actually inside. It's just to clearly keep things distinct. Otherwise, we would underrepresent our power, thermal, and the other parts of the business.
Thanks a lot. Just one more on just capital deployment. You said the pace of acquisitions has picked up. Some of your peers, their acquisition size has picked up as well. What's the probability of you doing a larger acquisition? Maybe help us understand any direction.
Yeah. I will talk in terms of, I go back to what we said, our pipeline is vibrant. We feel confident in our ability to execute everything M&A, from the acquisition itself to the integration. Certainly, as Craig explained, we have the wherewithal. With that in mind, let's see what the future brings.
Thank you. Just a question on bottlenecks. I think, Craig, you sort of addressed it's one of your top priorities. As you get geared up for 800-V, how should we think about potential bottlenecks, and specifically, can you get enough semiconductors for 800-V architecture? Can you get supercapacitors? I would imagine commissioning would also become much more complex. How would you deal with potential labor bottlenecks? Thank you.
You want to go?
I think we have to vet the supply chain. We have to understand where the supply chain sits today. I think we've been partnering When we're going through the designs with Scott and the team, I think we're starting to look at what does it take to deploy at that level and understanding what we need to be able to deploy and to deliver the product. I think we go through a rigorous process of doing that design, that development, and understanding what suppliers that we'd be able to have and when we would be able to get that product and be able to deliver on that product. I think that is part of the stage gates that we would go through before we deliver the end solution.
Scott will walk you through it, but there's several different ways that you would deploy that 800 volt. Depending on which way you deploy those, you would have some of those architectures or some of those pieces of outputs that you'd have, and some of those you may have in other different solutions. I think what we would be thinking of is not just the holistic 800 V, but what's in the way it's going to be deployed so that we can ensure that with the customer, we have all the right products, we have the right design, we have the right production schedule, because I think it's going to come in different ways and shapes and forms. I don't know.
A couple of points. Thank you. I would add to the very tail of your question about commissioning and services capability. We feel very good about that.
Yeah.
Not only because of our presence and a very well-trained and ability to continue to train, because we also have today a very strong, especially in North America, where it all will begin, a very strong commissioning and services team that is doing medium voltage today. It's a large organization that is ready to be deployed. Again, another aspect is to keep in mind, Andrew Percoco, is we should not think about something snaps and all of a sudden from the traditional, Scott Armul will elaborate on that, from the traditional infrastructure, boom, new infrastructure. Do you remember when air cooling was that, 3 years ago?
Yeah.
It's not.
Microsoft did, yeah.
That transition will happen. That transition will happen at a key stack. That transition will happen as one of the possible configurations. It will be gradual. We feel pretty good about our ability and relationship with the suppliers to drive them with us.
Hey, just a question on maybe just integrated solutions, OneCore broadly, and just, I guess maybe a multi-part question. First, at this point, maybe how broadly are you attempting to sell that into the market? Do you have the ability to deploy it at scale yourself if the customer interest is there? I'm also just curious in terms of getting the customer to take more of what you offer, how you bridge across that. For example, I think Compass historically has only been a thermal customer. I think maybe they still are. How do you get them to, as much as they seem to love you, get them to buy into the powertrain and other related products? Thank you.
Well, there are two aspects to that, Jeff. Thank you for the question. Let's go to the first part of the question, is OneCore, how much can that be really the main product? I go back to my multiple layers. We don't think that will be the only product. It's another layer of value. Not all customers will have that type of business model. Some will have. For those that really value the engineering light, their end, all things that very established data center players may have, just as their legacy, their way to do business, some others don't. Different type of customers will have different type of needs, and some needs will be addressed exactly by OneCore. The speed that is common to all.
We expect a gradual pickup in terms of adoption, but there are other elements in the converged story that instead are just ubiquitous in terms across the entire spectrum. SmartRun is a way to deploy and fit out a white space that reduces months into weeks, single digits and single digits. That is really gaining traction pretty much across the board. Again, just like we were saying, it's never binary. You see a lot of adoption of infrastructure solutions of integration across the spectrum of the construction philosophy, design philosophy. Power modules and prefabricated powertrains are becoming more and more frequent.
Think about the spectrum and when it comes to capacity, we have pretty much a global footprint when it comes to infrastructure solutions and have kind of the tip of the spear being OneCore, and that is being expanded pretty much globally. You will see one of the three factories we have in South Carolina here for that. We are building capacity as we see the demand coming.
Good afternoon, folks. A question on mix. Just in broad strokes, how you think about mix playing into the margin expansion walk that you provided. You clearly identified services acceleration, converged infrastructure. I'm sure we're going to hear more from Scott tomorrow.
Yep.
Maybe you could talk a little bit about expected margin profile on some of the higher growth areas that you see.
The services does have a higher margin, but just the way that it would mix, it wouldn't overly mix in terms of the out years to be, I'd say, significant in terms of adding to the margin walk. It stays right about the same path that the OE side's growing as well. You don't really see a marginally different mix whenever you're going up through the path. It is growing, and it's growing faster than it is today, but it's not, in our model, overtaking it significantly where you'd see a margin expansion on that. When you talked about the different products, and we mentioned this a few times on the converged infrastructure. The converged infrastructure or the Vertiv OneCore Vertiv SmartRun, all the different pieces carry their own margin, right?
If there's different levels of those and they come together, they don't really dilute or accrete based on what we'd sell them externally. The glue or what you'd say the other portion of the converged infrastructure, we get it about what we'd say our normal margin rate is. We kind of hold that at our normal margin rate, so that in and of itself doesn't dilute us at all. If anything, it slightly accretes us.
Yeah, thanks. Over here, guys. I want to stick on service. I also have a question around service. Have you guys seen a measurable increase in attachment rates over the past several years? What is the expectation for that trajectory of service attachment rates in the five-year plan? I guess what is the true recurring revenue opportunity, if you had to size it as, I don't know, like a % of revenue or versus what that is today? Thank you.
Let me take a stab at it.
Yeah.
Thank you, Nicole. We're not specific about the attach rate as we have mentioned in some occasions, but we're explicit about the fact that we really like the direction in which our install base is going in terms of complexity. Complexity and attach rate are strongly correlated. Strongly correlated, we have seen that happen. In general, yes, our attach rates are strengthening.
When it comes to the longer term, where that exactly is going, I think clearly we have our plans, we have our targets, we have our actions, but it's not something that we would disclose. When it comes to the recurrent part of our services business, it was somewhere in the slide. I didn't dwell on it, but in the important segment, there is a 25%, 75% in our services mix. That 75% is the life cycle. Broadly speaking, the life cycle. The recurrence of that life cycle is certainly very, very high.
I would say there's other portions of our portfolio that are recurring revenue. You could call them recurring revenue. We've never put a number on it, when you have a long-term agreement with a customer that buys the same product over and over and over again, we would think of that as a recurring revenue if they had a contract with us. We've never put our arms around and stated that is a number that we have published, we believe that that relationship that we built with the customer in terms of the buy-through and the flow is something that we would think of as a recurring revenue.
Scott.
Hello, guys. Talk a little bit about digital. How much of it is table stakes versus kind of something you can actually get paid for? Can you put your digital on other people's stuff? Just a little color there.
I think, again, I would start it with the fact of it's a little bit of both. It's a little bit of table stakes, and it's a little bit of something you can get paid for, depending on how you deploy it and your customer. I always look at the digital portfolio as you saw Chris up here. He said, "I'm not even thinking of buying a piece of equipment without Vertiv Next Predict." That's where you want it. You want it to be at a point where it becomes the stickiness. It becomes the reason why you come to Vertiv. It becomes the reason why you want to have an attachment, and you want to re-up your service contract because you are assured of performance, you're assured of serviceability, you're assured of a customer that's going to be there and help you build your optimization plan behind that.
In the way that I say it's table stakes, it's table stakes in getting the customer to, I'd say, get with you in a relationship environment. Above and beyond that is where you, I think, add on to the services lifecycle on the back end. That's going to grow the optimization portion of it. That's going to grow, what I'd say, like the recurring revenue portion of it. Also the fix before fail, right? We would think a condition-based maintenance on the back of that. Really helping your customer solve by them never coming down, optimizing their equipment. That's where you're kind of sharing the mutual benefit. I don't know if you think of it any differently.
No, I don't think any different to that. Yes, we have a price tag, if you will, to the capability of the Next Predict example, but it needs to be viewed as more holistic. It's like the preference.
Yeah.
Is the total value that we deliver to the customer. The answer is really both. Can't just isolate it on and off. We believe that what we offer is pretty advanced. It is making a further difference to our delivery of service, to the reason why people choose Vertiv. It's not just the individual product, it could be the entire system and the ability to optimize and start to derive behavior at system level and infer from the system level if there is optimization or even risk, that you don't just look at the individual piece of a powertrain, thermal chain, but you look at it holistically. We were talking about optimizing the use of every electron.
During the life cycle of a system, a lot of electrons can start to go to waste simply because the system is not fine-tuned like it's fine-tuned day one, if you will. That's what a digital Vertiv Next Predict and optimization drive. There was another angle to your question, and it's the third party. We like to service our technology. That's the primary reason. Our technology from a service and monitoring standpoint is quite potent. In this moment, we stay very much focused on our installed base. It's very potent.
I know you don't want to talk about take rates, but I have to ask, what kind of penetration do you have with Next Predict?
We launched it in January.
Yep.
This year. We are seeing good traction with Next Predict. It's still early stage. Clearly, good traction, especially everything new, installed base.
Thank you for doing the investor event. You talked about capital allocation. I think even after accounting for tuck-in M&A, dividends, and buyback, there'd be about $20 billion less that's available for deployment. You said you could consider transformative M&A. Can you share a bit more on criteria larger acquisitions might need to meet in order for them to be something that Vertiv may execute on? Thanks.
Yeah. I think we look at it in terms of value creation for the company, that's where it starts. That's even the bolt-on M&As. We will go in and look at all different sizes and shapes and say, "Do we really believe that there's true value added, either on the technology side, on a market side, on a growth side, that we wouldn't be able to internally fund and be able to grow?" When you think of it, investing in Scott and investing in capacity, I say Scott, but technology, he's the nomenclature for technology for me. Investing in those spaces, the returns are faster and the returns can be higher.
If I can get there through an M&A faster or if they can get me there in a neighborhood that would be a higher return, by all means, it's something we're going to consider and take into the portfolio and look. We're looking for value, and that's the baseline understanding of it. We look in value in terms of what do they have that we may not be able to produce. If that's something we need to get today, and it's a speed at which we can get to the market faster, of course, it's something that we want to do. Why we have done a lot of the smaller ones is we see true value in those in terms of scalability.
The scalability of those and all the ones that we have done on that page, you can see the true scale on them and what they've created in terms of value for all the shareholders. It's really a delicate balance. There's not like a thou must be this or thou must be that. It is a value creation equation that we put down and we look at and we evaluate in a couple of different ways.
Yeah. The other point, we're talking about transformation. We got to be careful with the way we characterize that. We like Vertiv as it is.
Yes.
We like the direction in which we operate in. We like the space. It is size-wise, can be kind of a departure if and when a good and the right opportunity presents itself. We'll not be timid, provided that we see the full value creation. No transformation, but continued evolution and continue to deliver on our promise to be absolutely lead in the space.
Gio, just like a year and a half ago, I think you said you want to be prudent with your guide. If I look versus a year and a half ago, maybe just frame the headwinds that you talk about now versus then. Are things better or worse, would you say, in sort of the global supply chain? How do you think about that when you came out with your guide for the five-year CAGR?
For example, a year ago, year and a half ago, the question mark about power and power availability, though we were talking already in terms of there's now been a ceiling, there's been a pacing item. We're not as crisp and clear like they are today. What we started to see happening is definitely happening at scale. Especially in the American market, a lot of self-generation or hybrid ways of powering. I think almost verbatim, I was saying a problem, we see a lot of capital, we see a lot of ingenuity being deployed. We see that this ingenuity is starting to hit road in terms of power being made available. A big problem right now is capacity in field for traditional construction. That is a big constraint. Supply chain. Supply chain is never easy.
Growing at the speed the industry is growing is no walk in the park. It is continuously working the supply chain. Do I think that there is a plateau somewhere that we hit or ceiling? No. We would have represented that, of course, in our trajectory. It doesn't mean that, okay, we go out and we find all the components, be that power electronics, as the question Andrew was specifically asking. In general, that requires a lot of work and a lot of partnership with the suppliers. I talked about a partnership with the ecosystem, with our customers, but there is a lot of partnership with the supplier. There will be hiccups. Various bumps in the road, the trajectory is the trajectory. I really believe in that trajectory.
Maybe just a similar question on the waterfall margin chart versus a year and a half ago. I look commercial execution a little lower, productivity a little higher. I assume pricing power is not changing, you guys tell me. I would assume it might even be getting better. It's really just about inflation.
Yeah.
You kind of offset that with productivity.
Yeah. We say in the charts that it's a prudent guide. It's a prudent guide because we do like our position in the market and being able to commercially execute. We also know that it's an inflationary environment and tariffs change on us, as we've seen even throughout as early this year. Our reaction to that is always something we want to make sure that we stay ahead of. I'd say it's a guide that we've looked at prudently, and we understand where it means, and we understand what we're able to do from a pricing perspective. We also know that we really want to drive productivity to ensure that if something does happen in the pricing environment, we have the margin expansion that we expect.
Hey. I'd be interested to understand some of the differences between training and inference data centers, particularly around your opportunity, how they compare between the two, and how some of your different technologies around modular and OneCore play into that.
Very often it's hard to separate and know exactly if the infrastructure that we provide or others provide will be utilized for training or for inference. There are cases in which there is an absolutely clear mandate for training. In reality, when you build an infrastructure, you build an infrastructure probably with a life cycle of 15, 20 years. It's hard to know now that you will need that for training for 20 years. What we are seeing now, probably already for a couple of years, is a tendency to build fungible infrastructure in that respect that can do both. When we think in terms of an inference infrastructure, it's an infrastructure that will have degrees of redundancy in general higher than training. If you will, in many respects, a richer infrastructure.
Hence, a little bit more on the right side of the TAM spectrum that I was mentioning. Some people may be tempted to associate inference to more edge small data centers. Not necessarily the case. That will drive some, but edge happens a lot in large or very large data centers.
Hard to distinguish. Fungibility is central. We like inference in terms of what it means for Vertiv. When it comes to the portfolio, I would say that is less about inference or training. It's more about the size, the type of deployment. 3, if and when, and as we see some more edge data centers, you shouldn't think about edge data center as a small data center. You're probably talking about 50 MW-150 MW. Very large by all means. But now we are accustomed to think in terms of GW, like GW. So much. But that type of infrastructure is everything Vertiv OneCore and fully, let's say, integrated converged infrastructure. You will see it increments of 25 MW-50 MW, as Scott will explain today. That is well-suited to that type of applications, but that can scale much bigger. Yeah, we like the direction in which the industry is heading.
Thank you. It's Deane Dray with RBC. I've honestly lost count of how many times services has been focused today and the superpowers and so forth. It certainly begs the question about the challenges in scaling services. It's obviously labor intensive. I've seen references to the number of training academies that you have. You called out 320 service centers, which is What are the challenges today in building out services, both from the labor side but also on your systems to be able to enhance productivity?
I would say that scale itself is a challenge. It's a challenge for all, but we've been pretty successful in delivering the growth so far that you've seen in the numbers. I don't remember exactly, but probably we're 4,600 or thereabout 18 months ago. The growth that we are delivering in terms of headcount in the field is big. Technology matters. Technology matters a lot. If you think about, for example, we're adopting AI for optimization of our field schedule and everything lifecycle deployment, and that is generating net availability increase for our engineers. There is a recouping capacity, latent capacity, thanks to technology. We inject a lot of technology using the tools that we give our engineers to make sure that they have the right technical information at the tip of their finger when they are operating.
That is, again, enhancing the ability to rapidly bring people up to speed. It's a combination of things. Again, I think an interesting story, and you heard it from Chris Crosby, said, "Hey, PurgeRite." PurgeRite is a story of now scaling it at national level in the U.S. and is exactly leveraging the various already existing and quite mature services locations that we have presence, that offices that we have across the country. The footprint that we have today is one that is very, let's say, conducive to growth. If we didn't have it would not just be about the academies or training people or having the right technology. If you don't have the local supervisors, the local managers, et cetera, it's very difficult to scale.
Having said that, you have to find the right people, and you have to train them from the beginning. We're accelerating that. Never easy. I think we're doing a good job there.
I think it also, and just to add on, I think it also comes back to the flywheel when you think of optimization and digitization of Next Predict. What that does to help you scale is when you can do condition-based maintenance, you don't have to have somebody on call all the time ready to go. You do have to have some people, but if you know that you're going to be doing regular running maintenance, you know that you're going to be able to be having certain checks and walks, and it does that all for you in a digital background. It simplifies the way that you go to service market and the way that you think about servicing a data center. I think that underlying and underpinning portion is being a lot discounted when you think of what you have to do to go deploy.
When you get that in a data center, it understands and it can read how the equipment is actually operating, and you can fluctuate it up, fluctuate it down. You keep the heats in the right levels, you keep the fluids in the right levels. You can take out a lot of what I would call the running maintenance of it is and potentially stop failures from happening. All of that helps you scale, and also helps your customer be more efficient, and also helps you unlock different ways to create revenue.
All right. Over here, guys. A great presentation. Two questions from me. First is on the point regarding the TAM, the $75 billion. If I put it all together, in the last 4 quarters, we saw 25 GW leased in the U.S. alone. If I take your 3 million per megawatt, obviously 325, I'm no math guy, but that gets me to $75 billion. That assumes that no data center is leased international or nothing is self-built. From your perspective, is this just being conservative or is there something that you're trying to imply in the TAM math? That's the first question. The second one is you think about the reweighting of data center demand. We saw the unlock using natural gas on site. Lead times are extending there. Have you thought about extending either down further into the data center or up to the power generation lane, just given how much demand there is for on-site gen? Those are my two questions.
Sure. Well, we certainly can go more into the details of the market model. We think that $75 billion, given again, our current portfolio plus the extension, because we're including that extension and considering the mix, and again, it's the entire range. It's not exactly the 3.5 point. If you think about the entire range, we're not too far away. Yes, clearly the U.S. and the North America is by far the biggest market. Probably, your assessment of total gigawatt actually sold is a little bit higher than us. We are conservative. We are prudent in many respects. If it will be more, as you say, the more, the happier. Absolutely, we'll not shy away from that. When it comes to extension, we certainly like everything that is power management. We're not into power generation.
We partner a lot with the power generation people, and that partnership saw Caterpillar in one of the charts is an important partnership for us. Those partnerships, of course, influence a lot the way we design. We have reference infrastructure, sorry, reference designs to make sure that the infrastructure can be rapidly deployed around that. When it comes going kind of absolutely owning everything in the data center space downstream from power generation or downstream from the, let's say, grid, well, that's now our space, and we'll continue to be obsessive about owning that space.
I think when you think of power generation, it's also a different market than what we think of in terms of a pure play for a data center. Power generation does drive power into a data center, but it does a lot of other things as well, and it's a different market, and it's a different technology. When I say even just a different industry in the way you go about it, thinking of once that power gets past the point where it's just going to a data center, we love that because we know everything about that. We know all the infrastructure, the physics around that. Power generation for that is a small portion of what power generation does for the whole world. It's a bit of a different market when you think of it that way. You're kind of, I would say, in that world, really mixing 2 spaces for us.
I just had a question on the converged infrastructure concept. I'm wondering if that moves the industry more towards standard designs and away from customization, or am I just thinking about that the wrong way? I guess in the concept of every data center being a snowflake and you guys addressing inefficiencies at the seams, was that something over time that maybe suppliers benefited from in some way?
Sorry, can you say it again, the second part?
Just the idea of every data center being a snowflake and this helping to address inefficiencies at the seams you talked about. I'm just wondering if that, over time, maybe suppliers benefited from those inefficiencies in a way.
Sure. I think that requires a little bit of a double take and the complexity of the industry behind. Different players in the industry have different business models. If you take the big players, be them hyperscalers in self-built or the large co-locators, it's not true that each data center is a sort of snowflake. There are very strong professional operators that have cycles of data center products that are being deployed. These are people that have know-how, expertise, engineering, and that optimize to their exact business. That part is increasingly being prefabricated, made in factory. That's a trend that is common to all. There is a part of the industry where, and that's particularly true for the AI factory, that, let's say, is ideally optimized on a certain type of silicon.
Well, that optimization and that standardization can be done regardless of who then the deployer and the owner will be. That's when the standardization comes to play. That's where the, let's say, hyper-optimization of the data center is happening. Yes, where going back to the stack, that value stack that I was showing, that's when the design around a certain type of silicon will be optimized and hence deployed at scale and at speed across a number of different customers and players. There will always be the nuances of the individual customer and their business. When I look at the data center space, I see kind of a multifaceted space where increasing, let's say, convergence and increasing fabrication in factory manufacturing is going to increase. Different players will have different business models, will have different type of needs.
Everyone, in their own way, is driving towards the elimination of snowflakes. Snowflakes are costly, and snowflakes are a lot of waste. Will other players benefit from this standardization, if you will, or optimization, extreme optimization? Possibly. That's okay. It's not about being unique. It's about being ahead. It's about being ahead. Sometimes not being alone is an advantage.
Maybe a quick follow-up for Craig.
Yeah.
On the investments, could you guys talk about the investments that you're targeting annually? I think last investor day you talked about a number.
In terms of M&A or.
In the bridge for fixed investments for capacity, for R&D, et cetera.
I'd say for our CapEx, we have a 3%-4% item on that for the period. Every year is 3%-4%, probably heavier, closer to 4% in the front of it. Back end, closer to probably 3%. If you're thinking of R&D, it probably sits right around 4% from a revenue perspective is where we like to be on R&D.
I have 2 questions. For the first question regarding the operating margin. 27%, are we also considering the next architecture, for example, freedoms, et cetera, a bit more higher density? Is this any correlation from this with the operating margin? This is the first question. Second question is about, we do have more than $20 billion capital that might be able to deploy for the M&A. If we are focusing on the future M&A, are we focusing on the upcoming technologies like 800 V DC, SST, or we can just simply invest on some of the company that might be developing the new technology of the company. Are we necessary 100% own this company?
Yeah. Correct me if I'm wrong, but I think your first question was around mix and how mix would potentially impact the operating margin. Is that correct?
Yeah.
is what I understood it to be. When we think of the actual framework and the growth of the framework, we don't necessarily see a massive mix perspective hitting us, in terms of growth. We would say everything kind of moves together, and we don't imply that there's going to be a mix up or a mix down from a margin perspective. Now, we do expect services to grow close to the 20% CAGR, but we also expect OE to grow in that same kind of framework as well. We're not really partaking or saying that mix is going to drive any of that margin rate growth. We would rather, if that happened, it would be a good thing, and if it didn't, we would offset it, is the way we would think about it. We really want operational items that are going to drive that.
On your second question on the $20 billion of capital deployment that we potentially have available to us. When we're looking at different sets of assets that we want to go purchase, and I can also ask Gio too, we come back to that value equation. We like the portfolio where it is today. We like the expansiveness nature. You saw all the different items that we have offer and all the pieces of equipment that we were able to bring to market. It truly is, can we invest internally and get something to the market that we believe is growable and we believe is scalable? Or do we want to go get that asset externally and grow it as fast as we can?
That's why you'll see us invest in things as small as a couple million dollars up to $1 billion, because all of those have different levels of what we think is value return for us. That continues to go through the entire scope of, if there's something that's even bigger than $1 billion, we would look at that as well in the same lens. It's not about we need a particular technology today. It's about where in the portfolio do we see value that we know we could go get and that we could grow and that we could scale and would add to our overall value return. I don't really think we circle an area and say, "Absolutely have to go buy something here." We'll circle an area and say, "Can we develop a technology for that? Should we go inorganic for that? Should we think about it's not a space that we want to play in? We do that for sure. This has to be inorganic right now.
Yeah. Okay.
Hi, Toby Okwara from Morgan Stanley. Wanted to ask some more about the content opportunity and how you mentioned that Vertiv's portfolio content is higher than that three and a half range. Could you give a sense of how it differs from the industry average? Is there any difference with the current AC architecture and how that content would be in the DC architecture for a data center?
Yeah, I think it's hard to talk about industry average. Our peers not always have the same type of portfolio, so it would be hard for me. Portfolio is really, let's say that range is really characteristic of our portfolio evolution and our mix, and our geography mix, and market mix. I would struggle going down that path. What was the second part of your question? Sorry.
Any difference in content for the current AC versus DC?
Sorry about that. I think we will hear tomorrow about that in technical terms, but I'm sure that technical aspects will be very illustrative in terms of the degree of complexity that if the current and future parallel evolutions of the technologies will bring. Fact is that the running those loads is not getting any simpler. If anything, it's becoming more complex, and that complexity is driving the multiple ways in which that power will be delivered. Again, the obsession here behind is how can you maximize of a given contracted power? How can you maximize the number of electrons that go to the GPU? Complexity drives opportunity for us in service, in TAM, and we firmly believe that DC in the rack, 800 volt in the rack or all the entire infrastructure are certainly not contradicting that complexity drive value for Vertiv.
Thanks. I just had one question. Given how you've talked about Europe sort of turning a corner in the last couple of quarters. When we think about your long-term guidance, how is Europe sort of either overly influencing the market growth and also the margin expansion in the next few years? Few years. Thanks.
You're talking Europe specifically or any region?
No, just, yeah, EMEA as a whole.
EMEA as a whole. We like where EMEA is going. In terms of what we're projecting here, we don't break it out by region. What I will say is, going back to what we talked about in the end of the first quarter is, we've seen a strong tailwind there from a perspective of what we're seeing in the pipelines, what we're seeing in the activity in the market, and what we're seeing execute. It right now has a great growth opportunity for us, and we see that continuing to happen specifically in the second half of this year, it bouncing back, and we would expect that to continue to go out in the future years, especially with the activity in the pipeline that we're seeing. We expect EMEA to be a good growth engine for us, not just going into next year, but beyond that as well.
Yeah.
Hey, guys. Donald S. Porter, Winslow. The ability to take share and grow above market, can you just sort of dig into that? Do you attribute that to more the modular integrated approach? Are there other reasons? What kind of explains or drives the share gain opportunity?
I think it's the value creation for our customers, in any way, a very strong presence. Again, I want to make sure that that value creation happens at individual product level, at system level, at entire infrastructure level. It is across the board. Let's take the example of the dynamic nature of the AI loads, and take the example of our power converters. Our power converters are extremely good at their job. That's no converged solution. That is the product. Dare I say, very beautiful, but still a box.
Nice box.
Still box. The technology in that, which is, of course, I'm being silly now, it's almost octave time. Can I say that? The fact is that the technology that we put in our products systems is a winner, and our ability to scale and our service presence. I almost go back to my competitive advantages slide.
I think you also look at his concentric circles that he had up there, right? Every layer of that is another layer of value for us, but also for our customers. When you think of buying a point solution, the point solution is efficient, and the point solution is efficient for the job that it does. When you buy a system, it is efficient for the entire system. Back to his point of an orchestra, right? You have the best violinist in the world. You don't want him playing violin the entire time. You want him to play violin when he plays violin the right way, and you want the pianist to play piano when he plays it the right way, and when they all come together, it's a great symphony.
If you just have one piece operating at its most efficient point, you don't get system-level optimization. You get point-level optimization, and the end customer is not going to feel that. They're going to feel really good about one spot, but not about their entire system. If the goal is to get the most out of the chip, that means the system has to be optimized, not just every individual point. Again, when we're thinking about it that way, that goes back to those concentric circles, all the way to Vertiv Services. Whenever you're packaging a Vertiv Next Predict on that or an optimization on top of that's making that whole symphony. That's the, dare I say, the orchestra leader saying that's making sure everything's working together. That's how you get it when you're thinking of value add for a customer.
Just to follow up on capacity, you've added a lot of capacity in a lot of areas. Where do you need it most? Maybe you have a lot of projects going on, but if you were to bucket it, what are the biggest capacity additions you're adding?
Honestly, I think it's across the board where we see strength in all the different product lines. We see it in converged infrastructure, power, and thermal. We have adds across the product line. I'd say even in the regions, we see it across the regions. Most of the deployment's probably still Americas right now, but that doesn't mean it's all centered to America. We are bringing on a new factory in Asia. We are doing some capacity expansions in EMEA. Of course, in Americas, we definitely have seen capacity expansion. I wouldn't trigger point it saying it's one individual area. We see growth in all the areas, and it is backed up from what we've been able to put online.
Hi, it's Andrew Percoco with BNP Paribas. Over the next 1 to, call it 3 years, the amount of volume potential you have, it's hard not to see incremental margins really hitting more like a 35-plus %, which would imply by 2030, to get to your 27% target, would mean you're probably doing below.
Dave's laughing because he's.
I mean.
Can't listen. Shouldn't listen.
No, I just mean by 2030, if you're doing 27% margins, that means that later the second, your incrementals are dipping below 30%, which seems also unlikely. Can you talk about the linearity or the pace of that incremental margin through the end of the decade?
Yeah, it's pretty linear. I point to the fact of when we think of the pricing out in the future and we think of some of the delivery on the productivity, we would say it's prudent and it's offsetting the unknowns that we don't know today. We didn't know the tariffs were going to change on us this year. We didn't know necessarily that the inflation was going to go up or that the war was going to happen. For all of us, it's making sure that we can offset any potential unknown that we don't see in the future. Now, if all that stuff doesn't happen and we produce everything, by all means, I'm right with you.
I want to ensure that we have a level of not just goal, but push for anything that could happen and that we are protecting ourselves so we can show the margin growth. That's the way that we're thinking about it and the way we're driving it. Dave's with you. I know he yells at me about it a lot.
Sure. The majority of the increase in the market growth rate was in cloud and colo, and didn't go unnoticed in their 10K, there was a new customer segment, which is weird. Brand new customer segment in a very established industry, and it was neoclouds. What portion of those 8 points, faster CAGR in cloud and colo, is directly attributable to new players, neoclouds, new entrants to this that aren't AWS, Microsoft, or Google? Thank you.
I think we go into details of the model that I would not be prepared to disclose or go into the exact details. Let me stay generic. When we talk about colo, cloud, we talk hyperscale, we talk colocation, we talk colocation for enterprise, we talk neoclouds. We think when we talk about those players, that there are exact boundaries. The gray areas are absolutely predominant. You see new clouds that are serving enterprise, at times that are serving some hyperscalers. It's really blurred.
Fungible.
The good thing is that we see quite a strong demand across the board. I'm not sure I can help you in those exact details, David, but maybe someone can. Maybe.
All right. Thanks for the presentation. Just wanted to ask on the converged solutions and how the growth rate for converged solutions compares to your overall business right now, and how you anticipate that evolving as you move forward here. Just curious, is there a notable difference in terms of the backlog conversion for a traditional order versus converged solutions such as OneCore?
Well, there are different dynamics in different parts of our, let's say, pipeline and commercial conversion dynamics. I wouldn't be going into details. We like the type of conversions and commercial conversions that we see with this type of offering. Again, it's not for all. That means that you have to be targeted to the part of the market that understands that, where we see that need and address it specifically. Again, no two converged solutions are created equal. A OneCore value proposition and, let's say, spread of dissemination in the install base is different from a SmartRun. It's hard to, and probably I wouldn't, even if it were easy, give you an exact answer, but we like the conversion trajectory. We like also the conversion trajectory on our more traditional business. That protects. That tap will go to market very carefully.
The easiest probably way to answer it, in terms of just not even numerical, but a way to look at it trend-wise is, of course, we saw an uptick in pipeline and things towards the end of the year, as you saw in our orders at the end of the year. They spiked up, and some of that was from an uplift in infrastructure solutions. What I would say there is the pipeline and growth is in line with the deployment being something that's happening more often now because of more complexity in the field and because of the need for it. Does that mean it's going to continue that trend? We hope. We like it. We think that's a really good way for the customer to get better outcomes in a quicker time to token.
Again, we don't have a crystal ball in the way that people are going to deploy either. We're prudent in it. We like our solution. We like our point solutions, but we think system levels and converged infrastructures are ways of the future, and we're investing in that.
All right. We'll take this as the last question.
Oh, geez. Now I have pressure. Oh, man. Pressure.
Brian Drab with William Blair. You talked about, in the slides, 20-35 gigawatts expected coming online annually. I think every call that I do on Vertiv or any stock that has exposure to this data center theme, people are trying to figure out when is it all going to decelerate. You used the word accelerate many times today. How do you view that in the model, and what's the source for that? Is that a management estimate, the 20-35 gigawatts? Because I see a subscript 5, but I don't see, at the bottom of the slide, I don't see what the 5 is referencing, by the way.
That's just strange. There should be a reference if there is.
I didn't mean to be.
If there is a number, but anyway, we'll check that.
Well, how does it ramp or when?
Well, when we say 20 to 35, it's never an exact science.
Yeah.
It's certainly never an exact science. Think about 20, let's say, today, and 35 towards the back end. We have estimates. We have market analysis. We compound that with our assessment, with the conversation with our customers, with the industry in general. Of course, the further out, the harder to have the exact projection. We think that we are pretty reliable and accurate, but who knows? I mean, so many things can happen. When it comes to is there a normalization somewhere? Maybe. Sometimes normalization, whatever normalization will mean in a market where the demand for AI compute capacity is huge, and the whole industry is just at its very infancy. Don't forget that to date, the majority of the use of AI is AI to human, let alone machine to machine to physical AI.
The hunger for capacity today and in the future, I think, is pretty convincing. What the long term future will mean, we don't know. The demand and the intensity of use of AI and the data traffic is not going to abate.
Okay. Thanks very much.
All right. Thanks, everybody. A very robust Q&A. Super excited. Lots of exciting things going on here at Vertiv. I would ask folks in the room stay put for a minute. We will be closing off the webcast.