ZDNet Executive Editor David Berlind recently sat down with Doug Erwin, chairman and CEO of RLX.

ZDNet: Thanks for coming to meet with us. What's new?

Erwin: This isn't just about why we will hang on, but why we will continue to be a big player. I know where IBM is strong, where HP is strong. I know what they do well and I know what they do not do so well. IBM was the second investor in RLX. When IBM saw how disruptive to their servers we were going to be, they had to decide whether to buy us or build their own. They tried to buy us. We didn't sell. So they built their own.

So, there's good news and bad news there. By IBM getting into this in a big way, we no longer have to be evangelists to what blades are all about. Blades are a hot commodity. The bad news is that we're up against IBM. Actually, that's not necessarily bad news. IBM has proven year over year that they are a very good technology company, they sell very good boxes, and are moving into the service arena. But, what they have failed to do time and time again is sell software.

Yes, RLX did invent the blade. But there's something else we invented, and that was the software to drive the blades into this modular computing arena, this compute on demand, this grid arena, whatever you want to call it. Big companies [and] little companies are all looking for ways to build complex computer systems with standard building blocks. They want the glue and the mortar to run these systems. Open systems have been a delight for manufacturers but a pain for users. All these OSes and databases and everything have not been the easiest for corporations to have a very simple system.

Back in the mainframe days, where you had one operating system and one door into the computer system, everything was simplified and very easy to protect. Open systems really changed everything. When I was at BMC, my team was responsible for bringing to market a product called Patrol. We were looking at a company called Tivoli, but they didn't have any working code and wanted way to much money for themselves. We bought a small company called Patrol and moved them from Australia to Houston, Texas, and created a market called "systems management." What was systems management? It was nothing but a single software tool that--with a single console--allowed you to managed a heterogeneous environment. After four years at BMC and some tremendous growth in our stock and market value, I left and created another startup.

ZDNet: When was this?

Erwin: This was the 1994-1996 timeframe.

ZDNet: IBM didn't have much, but Computer Associates was a player by then, weren't they?

Erwin: By the time I left in 1997, there were four players. There was CA with Unicenter, HP with Openview, IBM with Tivoli, BMC with Patrol. Those were really the four key players. All the other guys basically fell apart or were bought up. I left and started a small company called Pentasafe Security. It was a single console managing intrusions, audits, those kinds of things--the same things that systems management was, but it was security management for a heterogeneous environment.

When I joined RLX a year ago, I had been an early investor in an early round with a VC fund that I'm in. So, I did own [a share of] the company, but it wasn't until last May that I actually bought a large percentage of the company and joined after seeing two things. First, the hardware technology and second, just as important, the software that allowed us to manage this compute on demand brick, this blade. Just by way of background, prior to joining RLX, Scott Farrand was author of a product at Compaq called Insight Manager. You know it. It's very prevalent and one of the most well-used products. Scott knows what he did right there and he also knows what he left out. So, we're building on top of that knowledge and we built a very robust software suite, if you will, that allows people to get the full benefit of dialing up compute needs on demand, re-allocating [hardware] in real time, automatic failover, control of clusters, and helping with the remapping of storage.

So there are about 12 modules in what we call Control Tower, which is the family name of our suite. What RLX is doing better than everybody else, including IBM, HP and all the other 2nd tier people, is that it has a very easy product to install, reinstall, image and update, easier than anybody else. We've been winning awards in last the 12 months all over the world for best management software.

When we go head to head with the big boys, we win because we have integrated software into the hardware, and it's just easier to do. For example, IDC says that you need at least one full time equivalent person for every 50 to 60 servers. Now, I don't know where IDC gets those numbers, but I don't know anybody that has the equivalent of one full time person for every 50 to 60 servers. It's probably more like 100. We have customers managing 800 blades with one full time equivalent. Not just managing them, but reprovisioning them three or four times a day. You can reprovision one or a hundred of these in a matter of minutes. The installability of these things, that's a big thing. IBM says well maybe we don't have easy installing tools. But how often do you have to install? That's the very fundamental reason why blades were created. The compute module of this compute on demand, you're installing and reinstalling like some of our customers like three or four times a day. It's not just once when you buy the product. You've got to re-image, you've got to re-install, you've got to remap. These are things that the software we created in Control Tower allows us to do. We're not only going to be a viable player, but we're going to be around for a while, taking it to the next step.

ZDNet: When I talk about being around for a while, I don't think about the company going out of business. When I think of a company being knocked out of a tier, I think about a company like Sequent. They didn't get knocked out. They got acquired. When I say you won't last, I'm saying you won't last alone. You'll get bought. If you have a good customer list, chances are you'll get bought for your customer list just as much as you'll get bought for your technology.

Erwin: If someone wants to come and offer us a couple billion dollars for the company, we probably would talk to them.

ZDNet: Is that all it takes?

Erwin: Today.

ZDNet: Maybe Sun's Jonathan Schwartz will call you. I hear he wants to be in the [Intel compatible blade] business in a big way. You just have to start using Opteron.

Erwin: Sun and Dell, for example, are really not in this space.

ZDNet: They're not in it from an Intel hardware perspective, but Sun is certainly in it from a software perspective. When you look at their N1 stuff, they are very focused on blade computing, but I think it's primarily driven out of their SPARC business, and not necessarily out of their x86 business. But my expectation is that they will make those same tools available on the now-evolving Sun x86 story.

Erwin: One of [our advantages is the size of our installed base]. When you have a huge installed base like IBM has with its Director product or that HP has with Insight Manager, you have to be very careful that every change you make, every function you add to it, goes backwards and forwards. That's one of the issues that IBM has always had with software solutions. That their installed base is so huge, they don't fully understand how to (or want to) make the modifications and put the money into doing that. Plus, their salesforce really doesn't know how to sell software. They're a box-oriented group.

We have the advantage that we don't have a huge installed base. So everything we're doing is not only building to work with our stuff, but we're building to the industry standards. One of our biggest competitors is IBM. But one of our other biggest competitors is the 1U. It's the person who has more money to buy compute and the first thing they do before they buy is they go out and they look at the cool technology. They see blades and they say "That's cool, let's go check it out." And they go and they look at they say "It's not the form factor that's the problem, it's the cost. It's more expensive than I want. I'm going back to the 1U."

We were losing more business to people going back to 1Us. There are four or five major manufacturers in the world who produce 1Us. [So we asked ourselves], "Why don't we OEM a 1U?" So that when the customer says, "we don' t want to go buy a blade, we want to buy a 1U," our response is "Great. Buy it from us." Here's the same 1U you can buy from 20,000 other people, but here's the difference. We wrote a little card that you can stick in the back of the 1U -- you don't have to buy it today -- but you can tomorrow or next year that all of a sudden makes this 1U light up and talk to our Control Tower. Now instead of buying dumb terminals or dumb compute modules, they can begin to move into this compute-on-demand world and hopefully use our software to manage this and, oh by the way, you can begin to adopt blades as your future form factor.

So, we're building to standards. Not only for that reason, but because our Control Tower supports Pre-Boot Execution Environment (PXE), Intelligent Platform Management Interface (IPMI), and Simple Network Management Protocol (SNMP), we can manage IBM BladeCenters and HP Blades because they work off those standards. So, you can look at our software as really the trojan horse to get into accounts and then, as we're about ready to tell you, provisioning is not all we do.

Provisioning is about as overused a word as the word "partner." When IBM talks about provisioning, they really mean full provisioning. When HP talks about it, they really don't mean that [either]. We think our solution -- where we talk about imaging, installing, updating, tasking, repurposing all at the click of the button or a mouse drag -- is provisioning. What we've now got are customers beginning to say "Let's take it to the next step and automate things." We've got rules in our business that say "If this and that, then do this." We have not had the opporutnity for non-manual intervention for hot-swapping or failover. We'd send them an alert just like everybody else and then it's easy to do -- you point and mouse click and you reprovision another blade or a spare. But now, with the 6G announcement in July, we can put all of that in the background with automation.

You can write to me at david.berlind@cnet.com. If you're looking for my commentaries on other IT topics, check the archives.

You can write to me at dan.farber@cnet.com. If you're looking for my commentaries on other IT topics, check the archives.