“Think about the future, Eckhardt!” the hoodlum Jack Napier snarls at a crooked cop in the 1989 movie Batman. It was good advice – big changes were about to come to Gotham City resulting from Napier’s transformation into The Joker, one of the most grotesque villains of all time portrayed brilliantly on the screen by Jack Nicholson.
It is also good advice for professionals working in enterprise IT. Technology, like time, brings a constant stream of changes that we must come to terms with so our enterprises can adapt to new realities rattling the IT landscape.
How do we think about the future when we don’t know what it will bring? One good strategy is to gravitate to infrastructure that is designed to accommodate change. That category includes mainframe system design, which was conceived from the very beginning as an “anticipatory” architecture. Its designers and engineers set out to build a general purpose system for commercial enterprises that would have a much longer lifespan than the specialized systems at the time. They had no way of knowing what new workloads and technologies would emerge over subsequent decades, so they built in plenty of hooks that would allow integration of new capabilities into the core design.
So it is puzzling to hear some people say the path to the future now means organizations should “get off the mainframe” because it is “old technology.” Makers of x86 distributed systems might be forgiven for this thinking because they envy the mainframe’s preeminence in enterprise computing and they hate the fact that, despite their best efforts, the lion’s share of commercial transactions and data continue to be handled by mainframe systems.
It’s downright puzzling that otherwise well-informed IT leaders do not dismiss the idea that mainframe is “old technology.”
Daimler-Benz produced its first automobile in 1894. Today in 2017, when someone buys a $96,600 S-Class Mercedes-Benz sedan, are they paying for “old” technology?
General Electric began building railroad locomotives in 1913. Today in 2017, when BNSF or CSX orders a $2.55 million GE Evolution Series locomotive, are they paying for “old” technology?
The answer is obviously “no.” Why? Because engineers at these companies have continuously renewed the original product design with new technology to deliver true value, and customers affirm the value of the new technology with their orders.
In every other industry than IT, legacy is a positive attribute. The Porsche 911 first hit the streets in 1963. Since then the car has undergone continuous development while remaining true to its basic concept: a two-door, 2+2 high performance sports car, “as beautiful now as it was on day one. “
IBM introduced the first mainframe system in 1964. Today in 2017, when a global bank deploys an IBM z13, are they paying for “old” technology? No. Continuous innovation has pushed the boundaries on technology to new heights with every new mainframe generation. In the late 1960s mainframe memory was measure in single-digit megabytes, double-digit MIPS (millions of instructions per second) and clock speeds above 16Mhz. In 2017, mainframe memory is measured in terabytes, MIPS exceed 110,000 and clock speed is 5 Ghz, on a footprint that is a small fraction of the early models.
The mainframe remains the most important computing system in the world because the people behind it continue to evolve the technology to meet new challenges, and that will continue to be the case for years to come. As a new business era unfolds with the growth in cloud, advances in cognitive technology and new solutions like blockchain emerging, the mainframe’s anticipatory design and unique ability to accommodate change means it will play a leading role in the global economy.
Every customer that buys a mainframe today can afford to make other choices; instead they choose the mainframe because it has a winning formula: new technology that provides the best business value.
Editors Note: This post was written by Bruce Smith who is part of the z Systems marketing team, so I can’t take credit, much as I would like to for this post. This thought of the mainframe as ‘old’ technology continually irks me, so I didn’t hesitate when Bruce sent me this copy… I have also seen a good blog on CBR Online recently that speaks to the same thread, so you may want to check this out.