Not long ago, IBMs server business was a fragmented collection of fiefdoms, the disparate product lines running in parallel with one another. The result was a server business that was beginning to tank. But over the past few years, IBM has stormed forth, gaining market share and becoming arguably the top server company in the world.
Industry observers credit William Zeitler, senior vice president and group executive of IBMs Systems and Technology Group, with pushing the turnaround. Zeitler, who was the driving force behind the server groups big Linux push, recently spoke with eWEEK Senior Editor Jeffrey Burt about the future of the Armonk, N.Y., companys server division.
How did you get IBM back into the server game?
We started a new strategy a little over three years ago with the launch of the IBM eServer, but we actually began the investments a couple of years before that. And we have made considerable progress since that time. … [Since] the end of the third quarter in 2000 … weve gained every quarter. And very importantly, weve gained back more in the three years than we lost in the decades of the 90s.
[Hewlett-Packard Co. has] been losing pretty steadily during this period. Sun [Microsystems Inc.] went up for the first couple of quarters, and then in the beginning of 2001, theyve been dropping pretty considerably. The only other one making progress is Dell [Inc.], and Dell and IBM have been gaining over the last three years. Were the only ones making money.
The real story is, we have a strategy that is based on innovation and positioning ourselves in places where we can add value to customers. Dells got a strategy based on superior execution and a good business model. The two of us have been gaining at the expense of HP [and] Sun. And you can see the consolidation in the industry [on the analyst market share charts].
Can you compare IBMs server strategy with that of your largest competitors, particularly given Suns embrace of Linux?
There are three elements to the strategy that we put in place. First is around leveraging leadership technologies. We were always strong in high-end systems—mainframes, supercomputers—and the first fundamental decision was to leverage the technology we had in those systems to deliver leading-edge capabilities into the larger segments of the market, Unix and Intel [Corp.] servers in particular.
To do this, we reorganized all of development. Before we had mainframes, Unix and midmarket servers, Intel servers. We put them all together. … So that gave us new high-end Unix systems, that gave us leadership in Intel systems. We took scaling technologies we had in these higher-end systems and put them into new chipsets that worked up to four-, eight-, 16-, 32-way Intel servers, new system control stuff that created the BladeCenter system. We had continued our technology leadership investments into new microprocessors, new packages. Weve got a very robust next-generation family coming out [this] year across the whole line.
The second major element of this strategy was to very strongly support the open movement in all of our platforms—Linux, Apache, XML, Java. Our logic was that new applications, new workloads would come on those platforms and that we should have excellent capabilities for Linux, whether you wanted to run it on a mainframe server or an Intel cluster. What differentiates us here from our competitors is that we have been early proponents of this and consistent proponents of this. Sun, while they now claim support for Linux, for a long time was trying to protect the Solaris franchise. A lot of [Linuxs] success is substituting against traditional RISC Unix platforms. If you looked at whats really happened underneath market share shifts over the last few years, [theres] pretty dramatic erosion in the low end of the RISC-Unix space onto four-way systems, substituted by Linux on Intel or Linux on virtual partitions or things like that.
The third major thing we have been trying to do is introduce a set of tools and capabilities that make it easier for customers to manage these multitiered, heterogeneous environments that typify how they operate. Some of this has to do with system control technology built into the product, some of it has to do with autonomic capabilities we built into the products. Same idea: take things we had in the mainframes and high-end supercomputers and apply that kind of across-the-board for all of these products.
What differentiates us vis-a-vis Dell is more in the first and third areas. I think weve innovated. One consequence of that is a much stronger position in the high end. What differentiates us against Sun is mostly in the second area. We have been a more consistent supporter of Linux and the technologies in the open movement and that has helped us against Sun. What differentiates us against HP is execution. Weve just done a better job of not just executing against our strategy but executing in the field.