WRQ Sponsor Message

July 2003

The New Compaq and The Old IBM

(Continued)

The problem is not that HP are bad people or that they don’t make competitive products. In fact, the HP 9000 product line is highly competitive, both in technology and TCO. I personally managed an SAP implementation this year that included an rp8400 MC/ServiceGuard cluster of three vPars per node, with automatic workload balancing and resource allocation among vPars. We budgeted six weeks and compressed the schedule to implement in four, including a migration from a service bureau. Everything worked as advertised from day one, and the customer is ecstatic enough to sponsor a success story as well as deliver a paper at this year’s HP World.

And HP also has a good story regarding their work with Linux, which they’re not shy about promoting. Then there’s the upcoming Madison line of microprocessors and continued success in storage networking. Oh, and of course, industry leadership in the OpenView space. While HP may not be the patent machine IBM is, they continue to hold their own in turning research into great products.

Finally, HP is quite aggressive in pricing and earning market share. IBM still has the arrogance that prevents them from competing aggressively until they’re on the brink of losing. I guess a price war is just a little too vulgar for those with blue blood!

But what differentiates “The New HP” from IBM is the unmistakable impression that it’s all about HP and not the customer. All we hear about is HP bringing its costs down, improving profit margins and other bottom-line preoccupations. What’s lost in all this is how these initiatives affect customers. And even more important, what effect that has on the HP brand.

IBMers scoff at the HP “Invent” slogan. IBM, as I briefly mentioned, perennially outperforms all its competitors combined in the area of registered patents. So given that HP makes great products of no particular differentiation from its competitors, what does that mean? I believe it means that essentially HP has become a “me too” company with relatively low transfer costs for its customers.

In other words, it costs no more to convert from a 3000 to an iSeries (or pSeries) than it does to convert to an HP 9000, and perhaps less. And given that the HP-UX is not binary compatible with AIX, Solaris or any flavor of Linux, the HP 9000 is by definition proprietary, and hence risky. What looks better to you, a proprietary Unix platform offered by a company known for end-of-life-ing platforms when it’s no longer cost-effective for them, or a platform that has been continuously supported and enhanced since the late 1970s and continues to be state-of-the-art?

As far back as my high school days I managed to avoid IBM. Back then, I would send my card deck off to the regional vocational school where it would be run on a Honeywell mainframe and returned a few days later. Then in college I worked on a CDC mainframe before moving on to PDP11 minicomputers. Early jobs were on Prime and finally the HP 3000, before the HP-UX and NetWare viruses invaded. So I have no particular affinity for IBM anything. Yet I couldn’t escape the feeling when immersed in IBM training and subsequent business partner events that today IBM walks the walk that HP only talks. There is no joy in this perception. Perhaps HP should engage a few of the barbers and taxi drivers who now manage strategy for IBM.

Scott Hirsh, former chairman of the SYSMAN Special Interest Group, has been managing data centers for over 22 years. Scott can be reached at 510.435.4529 or scott@acellc.com.