EMC Pulls the Pieces Together

 
 
By eweek  |  Posted 2004-06-21
 
 
 

EMC Pulls the Pieces Together


Chief executive Joe Tucci is transforming EMC Corp. from a successful storage hardware maker to a developer of software that manages the data sitting in those disk arrays. At the EMC Analyst Day event in New York this month, Tucci predicted his Hopkinton, Mass., company would reach $8.1 billion in revenue this year, with much of the new growth coming from the sale of software created by EMCs newly formed Software Group, which combines the Documentum and Legato software businesses that EMC acquired last year. Tucci said at the event that server virtualization software developer VMware Inc., which EMC purchased early this year, will continue to operate independently. eWEEK Editor in Chief Eric Lundquist and Senior Writer Brian Fonseca discussed EMCs future with Tucci at the event.

Are you trying to take on too much? How do you balance off software and solutions with making sure you keep the hardware stream coming?

First of all, the world loves labels, right? We obviously have a hardware business, which is run by Dave Donatelli. We have a software business, [and] we have a services business. Software and services is about 52 percent of our business, and hardware is 48. Both are very important to us, and both are growing. Theyre all very much focused on a singular market: [managing] information. And VMware is a one-off—thats a phenomenal little company, Ill tell ya.

Hardware, software and services—sounds like IBM.

They have their systems business, they have their services business and then they have their software business. So its very similar. Theyre covering the whole waterfront, and Im saying were focusing around everything to do with information. Fixed content, unstructured content—if its about information, were interested, and were focusing on a much narrower area.

Do you feel you have all the pieces now and you just have to start revving it faster and making it work better together, or are there still some aspects you need?

I dont think there are holes or missing pieces, but there are always things you can add to make it that much better. I will never be satisfied because I think, once youre satisfied, maybe thats the beginning of the end. I think thats one of the things that happened to us back in 2001. We were satisfied, and we had convinced ourselves we had a two-year read [on the market] and then got caught by the competition. ... So youll never get me to use that word "satisfied," but we have the pieces to be successful.

One of the difficulties IBM had after acquiring lots of pieces is customers asking, "Whats the administration tool that sits over all of them? How do I know whats going on there?" Thats a chunk you still have to develop, isnt it?

We have a pretty good chunk there with ControlCenter (see story, right). Weve been ratcheting ControlCenter up over the years—this isnt something were starting from scratch. This is the framework that has what I call the male and female plugs. So you can take it and plug into a [Computer Associates International Inc.] Unicenter or [an IBM] Tivoli [systems management application]. On the other side, you can plug other pieces of software into that.

So some of these other [vendors] have had some of their own kind of monitors; what we have to do there is take that monitor and plug into the other side of ControlCenter. So you have all the plug-ins. So, for instance, a customer could have, say, an IBM array behind an EMC array. And all they have to do is say, "Replicate this data to here from there." And the way IBM does it, and the way EMC does it, is through totally different calls—and code bases and everything else, obviously. But the user wont care. Thats one of the things ControlCenter is set up to do.

Next Page: The problem of unstructured data.

Page 2


The biggest data management problem that a lot of companies have is with unstructured data. They want to get a handle on it and put it into a regulatory framework as well. Another problem is, How do you add security layers into the whole data management process? What is EMC doing in these areas?

We do it several ways today. Theres zoning ... and saying only that server or piece of that server can get that data. Its really a server function. The thing with storage is, youre working so fast, speed is so important. You tend to want to do that on the other side. But what I think we can do is when we [remotely replicate] data from here to, say, New Jersey, you may want that encrypted. We work with companies and third-party players to encrypt it while its going over the wire.

What are the limits of your agreements with Dell Inc.? Do you continue to develop low-end devices? Do you also push up into the higher end with Dell? Where are the borders?

There are no borders, really. Basically, its a matter of practicality in who does what best. Primarily now, with the CX300 and AX100 [storage arrays] and the follow-ons and offshoots of those products, weve probably covered a huge piece of Dells needs.

On the other side, when you get into big machines, youre going inside the data center, the way your servers change control, now youre getting into EMCs territory. Everything keeps getting redefined. The [management software] were talking about today is incredibly complex because its not just putting in the array, its putting in everything else. You put an array in, now youve got to replicate, you have to back up, you have to manage. ... Thats why were building out that big services organization.

A typical [data] life-cycle model goes through early childhood to early adolescence, maturity, old age, rapid decline, death. Where Dell likes to get [customers] is probably a little past adolescence. Like Dell today is one of the only companies out there without a tablet on the market. Its too early. They dont want to do all the R&D [research and development] and all that engineering. What they want to be able to do is say, "When these things mature, well take them to market."

A few years ago, there was a point of view that for EMC to dent the small and midsize business customer market it would need Dell. Now with the recent announcement of the NetWin network-attached storage device, it seems EMC is doing OK on its own.

Its really interesting. Dell accounts for one-third of our business. All the other partners combined account for one-third, and we do about one-third ourselves. But the one-third we do ourselves is a very different kind of sell. Thats where youll have a big chain of something that has a lot of remote offices that are still doing the central buy. We can service them like a large account, [provide] incremental storage units.

We dont have a way of getting to the small business. … The local K-12 school we dont reach. The big university, sure. Its the same thing with small legal firms, small professional services firms. [Smaller companies] are doing all the growing. Big companies havent created employment in forever. So [the Dell partnership] is great for us; Dell is still the biggest [partner] by far. Its a third by itself and all the others together equal Dell, which is good balance. But [EMC itself sells] mostly fewer [but] bigger things. So to get one [big] CX700 [array] you have to sell 20 AX100s or something like that? Our guys are out selling the one CX700 for $150,000, [while] Dells off selling 30 AX100s at 5K each. Theyre both $150,000 but a very different sell. While our dollar buy-in is the same, the product unit is not even close to being the same. So Dell is very important to us.

Its unusual for a CEO these days to make a real projection in front of security analysts with real numbers attached. There is so much ducking going on—its refreshing to see someone put up a real number.

Youre right, its either bold or stupid. Its really interesting. They want full disclosure, they want you to give good insight but they still want you to beat it. Well, if Im telling you everything, how am I going to beat it? Im telling you where I am. The best way to beat it is to pull back. So we decided to lay it out and well see what happens.

The opportunity is there. It absolutely is. But weve got to do a lot of work. Thats not a layup. But again, nobody is saying [that reaching our financial targets is] absolutely undoable because everybody knows you can do a lot more in the fourth quarter than you can do in any other quarter, and Q1 tends to be the smallest quarter of the year. It will be interesting to see the results.

Check out eWEEK.coms Storage Center at http://storage.eweek.com for the latest news, reviews and analysis on enterprise and business storage hardware and software.

Be sure to add our eWEEK.com developer and Web services news feed to your RSS newsreader or My Yahoo page

Rocket Fuel