Forget the consumer PC business. Dell Computer generates most of its dollars from enterprise hardware. Following Dell's launch of an Itanium 2 processor-based server, Kevin McKean and Mark Jones interviewed company Chairman and CEO Michael Dell about his enterprise ambitions.
Q: What are the strengths Dell brings to the enterprise, and what specific segments are you interested in?
Dell: We are the epitome of mass customization, that's what Dell does. In fact, I think we might have invented it for the computer industry. So when you talk about having highly customized, highly specialized computer systems that are configured exactly to the customer's requirement, I think we do that better than anybody in the industry. We have 15 percent of our business in the consumer business -- people … think that's Dell's business, (but) let's talk about the other 85 percent. We have, in the United States, more than 4,000 people that live and work face-to-face with our customers in the field. They're out there with Exxon and Mobil and Ford and Boeing and General Electric and United Technologies. What is characteristic of Dell's growth in the enterprise has been a shift from the proprietary computer companies that did everything … to more of an open, industry-standard model. You have the influences of Microsoft, Linux, Intel, Oracle, and many of the others collaborating. And we partner with (EDS), we partner with tons and tons of companies out there.
Who sells more servers in the United States than anyone? Dell. Who's sold more servers in the last two years than anyone? Dell. Whose server sales were 40 percent last quarter? Dell's. I think the inevitable forces (apply) here of leveraging high-volume economics. For some of the vendors that are still trying to live the old way, (I have) to tell you: I believe it's a high-volume, low-margin business. I don't think it's going to go back to the old way.
Q: Your server strategy is largely based on two- and four-way servers. How deep into the datacenter can Dell go?
Dell: I think first you start with (servers for) Web applications, those are easy (and) we've done those for years. Then you get to the core enterprise applications. You can look down the line from SAP, PeopleSoft, Siebel Systems, Microsoft, and others, and there's been a pretty huge embrace over the last couple of years. There were various debates and tug-of-wars about what's going to happen, but it's kind of gone (away). Then we (have) the database, which is another role, application, your sort of Holy Grail, a choke-point in the enterprise. Look at what Oracle's doing: It's (doing) scale out. I think you'll see SQL Server and DB2 and others going the same way. Scale out is going to make its way into the last vestiges of the mainframe. Not only (is scale out) massively lower cost, it (has) greater performance and it (has) higher availability, higher reliability. Essentially what you've got are these very low-cost building blocks.
Q: Your Itanium 2 processor machines are primarily targeted at high-performance clusters. Will Dell build architectures to go after the supercomputing space?
Dell: We've done some pretty massive installations, (such as) an installation at the University of Buffalo (in New York) that has 2,000 nodes. In the Intel world, according to some estimates we have upwards of 60, 65 percent share of the high-performance clustering market. So it's absolutely a focus. Are we going to go build our own 50,000-node supercomputer so we can show it off to the world? No. But if somebody wants us to, and they'll pay for it, we'd be happy.
Q: When it comes to the datacenter, there are some problems best addressed by a scale-up architecture and others best addressed by scale out. When does the scale-out strategy not work?
Dell: It doesn't work where the software hasn't been architected for scale out, which is all the old software. But that's more a question of where the industry was vs. where it's going. If you look at Oracle, they embrace scale out in a huge way, they do it themselves internally, they run on a couple thousand Dell servers, the whole business. I think you'll see more and more of the world leaning in that direction. SAP at Sapphire (its annual user conference in Orlando, Fla.) was talking about scale out. This idea is too obvious and too good not to take hold. The wish of the guys hanging onto scale up is (that) you've got this one application (and) you have to run it on this 64-processor, your (system) that eats raw meat for breakfast.
Q: You mentioned SAP. Where is Dell in terms of building out an enterprise applications business?
Dell: Who won the Pinnacle Award (given to partners) from SAP last year, the award for top partner? Dell.
Q: But is that purely on the hardware side?
Dell: No, I'm talking about applications. SAP installations. SAP installations on Dell products all over the world. There's a lot more testing collaboration when you get into these enterprise applications and a lot more involvement in terms of infrastructure services, professional services, (and) implementation services, which falls more into the (market of) Accenture and people like that. So we've been broadening out the relationships that we've been doing -- the Oracles, the PeopleSofts, the Siebels, and obviously the Microsoft relationship has been there for a long time.
Q: Most of those vendors are moving to build services-oriented architectures. What role does Dell play with partners when it comes to the future of enterprise applications?
Dell: There's a lot of discussion right now about systems management. Nobody's very happy about this stuff, and part of it is that you have this very disjointed set of architectures and convoluted and proprietary things. You have server systems management, you have application management, you have network management, you've got all this stuff (but) it's just not there yet. So we're in a lot of conversations with Microsoft and others about how you can tie these various systems management offers together, tie together the Dell/EMC relationship and systems management for storage into application and network and server management and client management. Because the cost of managing the infrastructure is, in many cases, quite a bit higher than the cost of the actual products.
Q: What's going to happen in the systems management space?
Dell: I think you're going to see consolidation, you're going to see customers increasingly forcing vendors to work together (by) saying, "I'm not going to accept this anymore." The challenge here is that so far the likelihood that you're going to have one solution for all customers is not very high.
Q: Dell is known as an innovator rather than an inventor. You leverage the R&D of your competitors. Do you believe there will always be a significant role for those inventors?
Dell: We invent things that save customers money and we have the highest return on our R&D of any company in the industry. We spend half a billion dollars a year in R&D and we've got 1,500 patents -- so R&D's quite important. But there are companies in our industry where if you invent something really, really neat, you're a hero. That's not true at Dell. You're only a hero at Dell if you save the customer money. And that could be innovation anywhere. It can be in R&D, it can be in manufacturing, it can be in service, in sales, it can be in process, it can be in the direct model, it can be at Dell.com. I think if you look at the actual results, you'd be hard-pressed to say Dell's not an innovative company.
There are primarily two types of inventions. There are inventions that help the customer and there are inventions that help the company itself. Let's say you have a printer and you got that printer from Company X and you're happy with your printer. All of a sudden, you need a new cartridge. You go to the store and you see 1,000 cartridges. Is that an innovation that helps you? No. Say you've got this printer and you've got three extra cartridges and you say, "I don't like my printer anymore, I'm going to get a new printer." So you get this new printer and these old cartridges don't work in your new printer. Does that innovation help you? No, the innovation helps the company that made the thing different. There's a lot of R&D in our industry that goes into protecting the proprietary profits of companies that is of absolutely no benefit whatsoever to the customer. It's a crime. These are things that customers are paying more (for) than they should. That's not innovation.
Q: What are your ambitions as far as 64-bit computing is concerned?
Dell: We don't really decide whether customers want to buy 64-bit computing, the customers decide that. We had an earlier version of Itanium, now with the (High Performance Computing Cluster) products we think we're ready for a higher-volume type of product. Now, the early volume we think will be high-performance computing. Anyway, you get Microsoft and others playing and you get more of a broad set of applications. I think Microsoft will have something to say about that in the relatively near future.
Q: What are the top three technologies that you're most interested in tracking over the long term?
Dell: I would say we're very focused on wireless. It is certainly a big theme across our business. Scale out is a big deal for us and all that means in terms of systems management and the enterprise market. And the digital home is a big focus for us right now.