Technology News Technology News

Another Perspective


There aren't many places where you can sell old real or virtual machinery for more than new, but it can be done in the computer business.  Legacy technology sold at a premium price prevails in IBM's proprietary i and z lines.  Still, just because IBM has worked its magic in the past and even though Big Blue promises to do it again in the immediate future, there really is no assurance of perpetual success.  Here are three and a half threats to the status quo: Amazon, Apple, and Google, and Wintel.

In the case of the IBM i platform, high priced legacy support is bound to the use of green screen applications and the relatively high price that IBM charges for the DB2 for i database on entry and midrange machines compared to alternatives.  Even when the 5250 type screens are scraped and hidden inside GUI-packing presentation layers, the underlying applications code and the body of middleware and systems software that support it much more closely resembles technology that was in its heyday a couple decades ago than recent software technology available on IBM's Power platforms.  A similar situation can be found in the mainframe world, where the high priced, general purpose z processors and associated software may provide support for venerable applications and mature (to put it politely) middleware.  This old stuff coexists with much more economical Linux systems that occupy identical hardware fenced off by firmware and lawyers.

It's possible--and I think likely--that IBM could not sustain either legacy business, and particularly the mainframe legacy portion, without the highly visible z/Linux offering.  Unfortunately for IBM, a key technological capability that distinguished z/Linux from rivals in the form of non-z servers, is losing its advantage.  Virtualization on X86 hardware is catching up with IBM's superb z offerings (and its excellent Power virtualization, too).  The way the X86 stuff eats into IBM's glass house z and i-on-Power opportunities is very much a typical disruptive technology scenario.  Customers can see the all-IBM offering is superior in many ways, but the X86 alternative, priced a lot lower than anything inside an IBM z or Power skin, is mighty tempting.  Whenever customers come to believe the X86 options can do the job and save a bunch of money at the same time, some of those opportunities will be picked off by Hewlett-Packard or Dell or perhaps Oracle.

Until now, IBM has been able to fight back.  Sure, there are situations where the business case for migration to X86 (and, often, another vendor) is compelling, but IBM generally wins a lot more often than it loses, so it keeps on growing.  And, because IBM's hardware pricing, particularly for the proprietary legacy part of glass house deals, is so high, Big Blue's shareholders can expect pleasant reports of growing revenue and profit pretty much every quarter.  When it comes to business performance, IBM is superb.

But current and emerging X86 systems are really excellent.  Also, Windows Server and Linux are very good, too.  Finally, middleware, including databases from IBM and Oracle and Microsoft along with a zillion other packages from a sea of ISVs is quite ready for strategic applications.

Even so, IBM could hold onto its base and expand into areas where Oracle, HP, or Dell have disappointed customers, which means plenty of shops, if computing itself wasn't changing so fast and so dramatically.

So far, IBM's most dangerous rivals--Amazon, Apple and Google--haven't attacked IBM at the vulnerable boundary between goods and services.  IBM's longstanding adversaries in the Wintel continent are too frightened by the AAG onslaught, the AAGravation, to seize the opportunity that has become hideously apparent, at least to IBM and maybe to everyone except the Wintellians.  That opportunity is simple: Drop a hundred gigabit portal into the glass house and use it for spectacular, clever, inventive services that sometimes supplement existing applications and sometimes supplant them.

Some of the smarter people at IBM, which means some of the very smartest people in computing, saw this coming some time ago.  That's why the two most recent lines of IBM mainframes offer cages that can hold blade servers and technology to integrate those servers with the big and big ticket legacy machines.  IBM knows that this solution is at best temporary, because it isn't just the blades in Blue cages that will keep the customers in line.  IBM has to figure out how to provide the nifty geolocation services, the front end software that plays well with smartphones and tablets, and access to the huge shared databases required to connect customers' specific applications to the huge libraries of data churned up by consumers and businesses as they traverse the digital terrain.

The offerings that may bleed legacy systems white, systems as small as an entry i and as large as a fortress sysplex, are not boxes that might be sold by HP or Dell or Oracle.  The stuff that can overcome IBM's best products and services is what you can already see if not in your office then perhaps in your private life.  You don't have to be enamored of Amazon shopping with the brand new Kindle or watching your 401(k) fade into dust with the help of the Schwab iPhone and iPad app or clogging your arteries with the help of the Five Guys Android app.  But I can tell you that if you are in the insurance business, it's gonna be the app (and what supports it) more than those mission critical batch runs that brings your company more customers.  And the same goes for the car business.  If you think there can be a future for BMW without end user apps (and all the stuff that it takes to support the apps), please stop what you are doing and go to Starbucks.  And when you are there, you can pay for your coffee with an official Starbucks app .  .  .  or if you are a more rebellious Android user, you can pay via the competition, an independent's superior coffee buying app.

What is going on is that Amazon and Apple and Google are not going to confine themselves to selling media and easily shipped merchandise and advertising.  Facebook and eBay are going to play hardball with AAG and in some ways they already are.  Right now you can use a smartphone to buy a Power Systems server on eBay.  There are software and services companies showing off on Facebook.

Okay, maybe nobody has yet come up with the right blend, with that friendly app that lets you hook up with an HR services company using your iPad or Android phone and, after due diligence, dump that hairball of HR applications you've been running for years.  Well, that's now and this is next week.  You know the HR software specialists (and companies that envy their opportunities) are all trying to puzzle out how to make a bit of Java on a thin client give life to their ambitions.

It might be that cloud computing can replace legacy applications on legacy platforms, but it turns out that this isn't really what has IBM worried.  IBM already offers that kind of remote service stuff and it hasn't wrecked its sales of i and z systems.  No, what has IBM nervous is the possibility that the legacy boxes stay put, maybe even grow a little from time to time, but that all the other computing, which is growing many times faster, will leave the glass house and disappear into the Googlesphere or Amazonery or Appledom.  Alternatively, if the Wintel systems vendors can overcome their bowel distress, perhaps it will be Dell That Remembers When Michael Was Smarter or The Company Formerly Known As Hewlett-Packard or The Sun That Oracle Ate And Hates will spot the billions of dollars and thousands of computer professionals who are currently hanging around the glass deathbed reverently called Legacy IT.

Of course IBM could possibly invent itself out of trouble and transform its i-z legacy necropolis into a lively, fun, creative fountain of IBM-based services.  But I don't think the suits at the top are likely to be moved to action without first enduring some very frightening episodes.  And right now they seem a little too pleased with themselves and their predictable earnings per share growth.  As if that is what really mattered.

— Hesh Wiener October 2011

Copyright © 1990-2019 Technology News of America Co Inc.  All rights reserved.