Editor’s note: This is the second in a four-part series about the IBM’s mainframe’s past, present and future. For more, see "A Brief History of the Mainframe," "Not Your Father's Mainframe" and "The Next Generation."
Over the years, as lower initial cost technologies—like personal computers, distributed UNIX systems and blade-based systems—have come to market, the industry has been questioning the future of IBM’s largest systems. In fact, the demise of IBM’s mainframe has been predicted since the ’90s, yet despite negative press, the platform continues to thrive, with new education initiatives, hybrid capabilities, green computing and a steadfast commitment that has lasted for 48 years. To quote Mark Twain, “Reports of my death have been greatly exaggerated.”
So where do the misconceptions come from? And why would anyone consider moving off a platform that has:
• Kept pace with all the technology changes over the past five decades
• Preserved software investments, and
• Continually improved price/performance with each announcement?
One reason comes from a surprising source, a 2004 college textbook, “Computer: A History of the Information Machine.” In it, authors Martin Campbell-Kelly and William Aspray take the reader through a history of IBM’s mainframe and accurately portray events up to January 1993. They write “within a period of five years, IBM was to suffer the most dramatic fall from grace in corporate history, recording even more slender profits, until in January it reported the largest loss ever by a private company.” They fail to mention IBM’s complete turnaround in the mid- to late-1990s, and beyond.
In April 1993, Lou Gertsner, former CEO of Nabisco, took over as CEO at IBM. Under his nine-year reign, IBM reinvented itself, eliminating the silos between organizations that had been stifling innovation and growth. Instead of dividing the organization into smaller companies, “He organized IBM around global teams focusing on a dozen major industries…. By shifting from a being a country-centric, product-focused company to a global, client-oriented organization, IBM was able to spot business problems sooner and respond more quickly to shifting customer needs,” according to “Making the World Work Better” by Kevin Maney, Steve Hamm and Jeffrey M. O’Brien.
“Without [Gertsner], I don’t think we would have survived. We needed somebody with that tough mind and analytical skill,” said his successor, IBM CEO Sam Palmisano, in Hamm, Maney and O’Brien’s book.
In their textbook, Campbell-Kelly and Aspray also wrote that: “The mainframe had matured and became a commodity, while IBM’s expertise in systems integration was increasingly taken over by software.” While it’s true that over the past three decades IBM has moved into new business areas, like software, consulting services, outsourcing services and financing, the mainframe has never been a commodity.
In the computer industry, a commodity is an item that has reached its lowest price point in the marketplace, and can easily be purchased from many vendors. However, former vendors, such as Burroughs, UNIVAC, NCR, Control Data, Honeywell, GE and RCA, have left the mainframe market. In fact, Intel, AMD and IBM are the only vendors in the industry that manufacture high-end processor chips, like the ones used in the IBM zEnterprise System.
Students reading this textbook will graduate, assuming that the mainframe is dead, and IBM is on the verge of collapse. Such misinformation would then, in turn, be brought into their working lives, potentially prejudicing future business decisions.
Another reason comes from AFCOM, an organization of data center management professionals. Each year, it conducts a member survey, asking questions relevant to the IT industry. In its 2009-2010 survey
of data center trends, with 436 of their member data center sites responding, AFCOM reported, “Approximately one-third (32.9 percent) of all existing mainframe data centers will no longer use mainframes in the future.” This study sent shockwaves through the industry.
Technology magazines, bloggers and others started writing about the mainframe’s impending demise. The interesting thing is, the study didn’t explained why clients would no longer be using mainframes in the future. Some possible reasons are reported in other parts of the same study. For instance, 62.1 percent of respondents indicated they are consolidating one or more data centers, with their number one concern being availability of sufficient power.
Lastly, the study revealed that 14.9 percent of their respondents were already using cloud computing, with another 46.3 percent considering it. Emerging technologies, like cloud computing, are weekly topics in many of the industry’s trade publications. In addition, server consolidation is a major part of many data center consolidation projects. These may have contributed to the survey responses about moving off the mainframe. However, these results are at odds with other recent studies in the industry.
A 2011 study
conducted by BMC Software, a large manufacturer of mainframe software, states that “93 percent of respondents at large companies expect [mainframe] capacity to grow or remain steady.… 47 percent of respondents said that new workloads and new business applications are contributing to their capacity and growth.”
The 2011 IBM zEnterprise System announcements may be contributing to such findings. The zEnterprise ecosystem is the latest move by IBM to embrace hybrid computing, allowing shops to run and control multiple system environments within a single machine footprint.
Another important aspect of the AFCOM study is how data centers are embracing cloud technology in their future plans. The next installment of this series will examine how IBM’s mainframe has not only embraced cloud technology, but also pioneered the technology. It will address how today’s System z platform runs diverse workloads, including UNIX, Linux and Windows, besides z/OS and z/VM.
Al Grega is a worldwide sales executive for IBM WebSphere Software on the System z platform. Previously, he owned worldwide sales for IBM Rational Enterprise Modernization software and compilers. He has more than 30 years of experience ranging from OS development to services, marketing and product management.