The enterprise datacenter isn't going away: Vertiv CEO Rob Johnson Credit: VertivSupplied Art
Business Management

The enterprise datacenter isn't going away: Vertiv CEO Rob Johnson

Vertiv calls itself ‘Architects of Continuity’ that collaborates with organizations and customers to envision and build future-ready infrastructures. There are many challenges facing today’s datacenters, communication networks and commercial and industrial facilities with the emergence of digital technologies like IoT, HCI, big data, 5G and more. 

IDG had an interaction with Rob Johnson, CEO, Vertiv, on the company’s extensive portfolio of power, cooling and IT infrastructure solutions and services - from the cloud to the edge of the network. 

Edited Excerpts

What have been your key priorities as the Vertiv CEO? What’s worked and what’s not? 

The main thing we’ve been working on is our company’s large investments around product development. We had colo (colocation) and hyper scale strategy to take market share which requires a different product than the enterprise requirement. 

We’ve had a channel/edge strategy that we’ve implemented, and we’ve developed products for both ends of the spectrum. Also, we did a couple of acquisitions to support colo and hyper scale and a couple more in the edge space. 

The colo and hyper scale, what we call named accounts and products, have worked really well. I would say the channel has been more difficult to get into because that’s more of a full strategy and there are well-established competitors there. However, we are making good progress. We are bundling our cooling and power offerings as a package system, and we’ve been able to execute some turnkey projects and utilize our global service organization. 

What key datacenter trends do you foresee in 2019 and 2020? According to you, how will Vertiv score ahead?

There’s a big rumor out there that enterprise datacenter is going away, which is absolutely untrue. In fact, our enterprise datacenter business in 2018 grew over six percent. We believe that with hyper-converged on-prem/off-prem we see a resurgence of the enterprise datacenters being rebuilt, driving up for higher density. Because the old traditional datacenters were designed for a low 2-3 kilowatt rack. We see the kilowatt rack climbing six to eight in the enterprise, and for the colo hyperscale it reaches up to ten and twelve. 

Over the last 18 months, the hyper scale team is not only building their own datacenter but splitting half their builds with colos. Several aggressive traditional colos (that wouldn’t play in that space) have really leaned down their design to actually house and help Amazon, Microsoft, and Google, amongst others. 

How are your customers buying critical infra solutions and services for datacenters? Any major changes there? 

Over the last 36 months, there’s been a change in hyperscale and specifically enlarged colos demanding the kinds of product they want and engineered accordingly. They used many local vendors who couldn’t scale globally, hence we came into play. When you look at an Amazon or a Google or a Microsoft or a large colo like Equinix or DLR, they want a global product set, more standardized products and they want an offering anywhere in the world. And that’s the new big change for us. Customized datacenter solutions for big organizations is an absolute given. 

The other uptake is around modular datacenters which are prefabricated, preskinned, tested in the factory as there’s just not enough resources in the field for the installation. The trend is not only on the hyperscale but also on the edge. 

And what happened to DCIM that anybody and everybody in the datacenter industry was gung-ho about till 2017 and 2018? 

As soon as Gartner took DCIM (Data Center Infrastructure Management) off its magic quadrant, you know what happened there. We’re still servicing, supporting, selling new installations of Vertiv Trellis platform (real-time infrastructure optimization platform) for somebody who wants to go deep.

It’s (DCIM) not emerged as the money-making industry that everybody thought it would be. Because instead of using one big unified product, IT managers are still using multiple tools for what they’re trying to manage. But, we still have very large installations of Trellis and we continue to add features to that product line. 

People want the data on the data lake, they want predictive analytics, and they really want to know when something’s broken. They don’t want a whole bunch of nuisance alarms. Our customers, especially on the edge side (we’re doing some very large retail, big box stores that are competing against Amazon), basically want to outsource the entire infrastructure to us. If something is broke, they don’t even want to know about it, they want us to monitor and take care of it. 

Who’s your chief influencer (or decision maker) at the customer end - CIO, datacenter manager or LOBs – that buys Vertiv’s portfolio?

At very large datacenters, it’s still more facility type oriented, engineering-oriented, but at closets and edge, it’s IT team and CIOs. Certainly, in the hyperscale side, they’ve built entire datacenter teams and that’s who we interface with. But, for example, we’ll work all the way at the top with Microsoft, because it’s important that they have a strategic partner who can scale and move quickly and do it pretty much anywhere in the world. 

So what’s your strategy to outdo your closest competition - Schneider Electric? 

I was with APC in the past, so I understand their strategy very well. On the small end, the thermal side, people are looking for more complete solutions. And that’s where Schneider really doesn’t have a global offering. They may have some pieces in regions like India, but we have global thermal offerings which is a big advantage as we are tying that with services.

We have over 3,500 field service people globally to help do turnkeys, to help that small, modular datacenter, that three rack datacenter, with the thermal management, with a generator back up, and help them maintain that throughout the life. That’s, I think, where we’re a little bit different. 

Would 5G and multi-cloud will be the new king in the world of digital transformation? 

For an infrastructure provider like us, where we do really well compared to our competitors, is our footprint in the telecom space with DC power and thermal offering. As it relates to the multi-cloud, there’s still some new entrants, but it’s getting harder because the systems are getting more complex, they want more of a unified system with monitoring predictive failure analysis, it’s harder for new entrants to get into the game. 

Edge is a journey which be greatly enabled by 5G. When will it be rolled out? What’s the business model for 5G? Who will pay for that infrastructure and services? There is a big classic edge in place with retail, healthcare, financial institutions, anything that has multi-branch network, physical footprint. It’s already happening. 

Dell announced datacenter-as-a-service. Do you see days of ‘everything as a service’ arrive? 

Absolutely. In the past, people had the mindset “I’ve got to own a datacenter and it’s mine.” Now, they want their workloads to perform, hence they really don’t care whether they own that asset or not. 

Big banks, especially in America, are selling off their big datacenters to real estate firms and others. It’s not necessarily popular to own that anymore, and I think that’s what Dell is looking at, and I like the approach. I know it’s not fully-baked, but nobody really wants to own it. People want to spend money on applications, they want to spend it on other areas. We’re in the early days of ‘datacenter-as-a-service’, but I believe it will catch on. 

What about ‘datacenter in a box’ because it’s a fairly loose term - as if stack everything up and it becomes data center in a box?  

For colo, it’s continuing to drive the design to be cost effective and increase the efficiency. And Emerson Network Power, now Vertiv, has always been strong on the thermal management side. That is, 60 percent of the power consumption in the data center. We continue to focus on layouts and the designs to drive higher efficiency and higher PUEs. 

On the modular side, we’ve just got, from a single rack, which is fully monitored, managed, all the way to a complex one. I think our largest is 30,000 to 40,000 square foot - all modular, including the offices and everything, flown in or brought in on big ships, and brought to sites in the world. 

Besides making it cost effective and efficient, speed is important. Organizations have trouble planning today, so the more we standardize, more we modularize, and more we can test done in the factory. To illustrate, if we can do a tier three commissioning at factory level it saves time in the field. Speed is everything than just about low cost. 

You spoke about Emerson Network Power which transitioned to Vertiv brand. Any brotherhood with Emerson wrt customer base, channel ecosystem, or it’s all separate?

It’s completely separate. Emerson still does have a small ownership in us but we’re fully independent. Strategies are completely different. We do some work with them in oil and gas on the industrial side, but for the most part, we’re separate and that’s the real reason why I like this spin out from Emerson. 

If one thing has changed in the past 24 months, it’s the speed at which we’re developing products, getting to customers, and coming up with innovative supply chain solutions. 

Any best practices for CIOs and datacenter managers to run efficient and scalable datacenters in digital transformation era? 

More standardization of their datacenter. The more they can standardize, the easier they can manage multiple deployments and multiple sites, whether it’s our equipment or somebody else’s. When people mix and match, it’s very hard to get the communication platforms, the monitoring platforms, to work together. 

They should deploy more open products that will allow for management of multiple people’s devices, not just only their own. I think the days of people having these “lock you in and you can only use my system that doesn’t integrate with others” doesn’t work very well. So, I’d look for more open platforms from a monitoring-management perspective. 

Your 2019 priorities as Vertiv CEO. 

We’ve been growing as a company much faster than we had expected. For us, the customer experience is the number one thing we’re really focused on in the end-to-end delivery. Making sure that, from a quality delivery, to a quality startup, and that whole bit - so really focused on supply chain and customer delivery. 

Also, people development is a focus as there’s shortage of datacenter knowledge-type people. Each year, we’re bringing in, hundreds of students out of college; we have internship programs that we we’ll run this summer. We are trying to help train and bring more people because we lose a lot of people to players such as Microsoft, Google, Dell, and others. In the end, we continue to deliver on our innovation and hyper-focus on channel and you’ll see us become more aggressive on the channel, and the edge. 

Too much data sprawl is goldmine for datacenter focused companies like Vertiv. 

Yes! More the data, the more somebody has to crunch it, more machine learning and AI. On the military side globally, drones are a big thing and they need local compute, permitted air zone, data collection and analysis in real time and to be able to act upon it. 

It’s all about the workloads and related aspects. Facial recognition is becoming real as cameras will drive a lot of trends from security, access standpoint, but also, don’t underestimate the autonomous vehicles. 

We manage infra for a lot of football or soccer stadiums and there’s huge amount of sophistication of computing and storage that’s going on their on-site for latency purposes during a game so they can deliver the right content to your phone, know your location, your profile, and where you want to go. So, there are so many cases right now for us to expect the next five to seven years will be really good. And Vertiv is that end-to-end infrastructure provider for all applications and all kinds of workloads of modern companies’ IT stack.

Rob Johnson’s Key Datacenter Takeaways

- The hyper scale teams are not only building their own DC but splitting half their builds with colos.

- Modular DCs - prefabricated, preskinned, factory-tested are popular as there aren’t enough resources for installation.

- DC standardization helps manage multiple deployments, on our equipment or somebody else’s.

- Deploy open platforms that allow monitoring-management of multiple people’s devices, not only their own.

- “Lock in to my system that doesn’t integrate with others” doesn’t work anymore in modern app world.

- “I want to own a DC” mindset has changed to ensure workloads perform irrespective of who owns asset.

 4 Priorities for 2019: Vertiv CEO

  1. Product Innovation: Innovate quicker as we are growing faster than expected.
  2. Customer Experience: Focus on end-to-end delivery, supply chain and customer delivery.
  3. People Management: Hiring DC knowledge-type people out of college through great internship programs.
  4. Channel Focus: Continue to be more hyper-focus and aggressive on the channel, and the edge in 2019.

PREVIOUS ARTICLE

« RBS cuts employee travel with collaboration technology

NEXT ARTICLE

Corsair's MP600 SSD will feature PCIe 4.0 to support 5GBps transfers »
author_image
IDG News Service

The IDG News Service is the world's leading daily source of global IT news, commentary and editorial resources. The News Service distributes content to IDG's more than 300 IT publications in more than 60 countries.

  • Mail

Recommended for You

Trump hits partial pause on Huawei ban, but 5G concerns persist

Phil Muncaster reports on China and beyond

FinancialForce profits from PSA investment

Martin Veitch's inside track on today’s tech trends

Future-proofing the Middle East

Keri Allan looks at the latest trends and technologies

Poll

Do you think your smartphone is making you a workaholic?