One on One: Paul Maritz, VMware Chief Executive

Paul Maritz at the EMC World conference in Las Vegas in May 2011. Ronda Churchill/Bloomberg NewsPaul Maritz at the EMC World conference in Las Vegas in May 2011.


Paul Maritz has been chief executive of VMware since July 2008. VMware’s server virtualization has made possible a vast consolidation of the computer business toward commodity chips and open-source software. It was also critical in the shift to “cloud” computing. A native of Rhodesia (now Zimbabwe), Mr. Maritz joined Intel in 1981. From 1986 to 2000 he was at Microsoft serving on its executive committee and overseeing Microsoft’s expansion into client-server corporate computing. He then founded a computing company which in 2008 was acquired by EMC, which also owns VMware. The following is an edited version of an interview with Mr. Maritz:

Q.

What is your most interesting business opportunity?

A.

Fifty percent of the world’s computer applications now run on virtualized servers. We are deliberately trying to make a transition with this company, from a server transformation proposition to overall information technology transformation.

Q.

You mean cloud computing. How will that change people’s work?

A.

Employees are no longer satisfied with a Windows desktop at work. The changes will affect not just the infrastructure companies put in place, but the employee experience, and the way a company relates to its customers. Most I.T. departments got good at managing a single customer experience around Windows. Now it’s a world with multiple devices, fixed and mobile.

The expectations customers have about consuming information – about products, their service, pricing, everything – is radically changing. Business, meantime, is sitting on 50 years of infrastructure that isn’t appropriate to the job. Their computing is built to send out paper bills – infrequent, narrow transactions. When the I.T. departments think of transformation, they usually think about putting that paper bill online, not delivering information with a greater context, in real time.

Q.

You sound a little like a marketing guy. Just how is the old technology wrong for the new era?

A.

There have been three big eras in computing: Mainframes, client-server, and now cloud. Each era has its canonical applications, its biggest drivers. With mainframes it was about bookkeeping, which rode on a database called ISAM. It still sits under most banking applications. Client-server had the Web, e-commerce, customer relationship management, E.R.P.; it was all on relational databases that could parse more complex relationships. Now there is cloud, with large scale computation, real-time information, and contexts. You couldn’t really do it on a relational database, and that is why Google, Facebook, Yahoo and others had to develop new databases like Hadoop, Cassandra, and Big Table.

When the data fabric changes, lots of other things change, too. You get a different user experience. The cloud customer experience only works if you upgrade your legacy applications, and use your own data in new ways. You have to pull in data from outside your own company to tell you things like where the customer is now, and what is located nearby.

We’ve been the remedial technology for client-server moving to big data centers. Now we have to become pivotal to this new kind of data flow. We do it by moving from selling just server virtualization to a suite of software to manage more of a company’s cloud.

Q.

Don’t you worry about consolidation? Some analysts see a time in the near future when just a handful of companies like I.B.M. and Microsoft will control global clouds that handle most of the business computing work.

A: I think there will be tens of thousands of clouds, because they will be so easy to set up. You will be able to buy or rent a portion of someone’s data center and configure it how you want. And cloud won’t be limited to megadata centers. There will be on-premise clouds that are cheap and easy to run. The really hard choices for most businesses will be figuring out which legacy applications they leave in existing computers, which they freeze, and which they move into the cloud.