IBM Systems Magazine, Mainframe - March/April 2009 - 26

How did you first get i nvolved w it h I BM a nd this project that’s grown over the years? Q: “The mainframe continues to evolve very substantially and has since its inception 45 years ago.” —George Walsh, System z vice president and chief technology ofﬁcer A: I joined IBM almost 44 years ago in New York City. It was just after the a n nou nce me nt of t he System/360. My original mission was help on the non-ma i n f r a me t h i ng s while people were being trained on mainframes. Ultimately, I took almost three years of education; that was the training required at the time to understand the new OS, the new architecture, the new systems. My first real mainframe customer was Con Edison power company in New York. It was amazing to see the things the mainframe could do for Con Ed. One of the very first things they did on the mainframe was essentially support customer service. That need grew into Customer Information Control System (CICS*), which is today the most used system, still running the most transactions in the world. It was a very interesting environment and a very innovative customer. Some lessons were learned along the way; the first systems weren’t as reliable as we wanted. There was a 24-hour outage and the customers were so mad they created a riot and actually burned down a Con Edison branch office. That instilled in me very early the importance of extreme reliability and availability and it stuck with me my whole career. What are some of the major changes or innovations you’ve seen in the mainframe over the years? Q: A: prevention prevent. When you’re looking at the blue screen, you should think of mainframe innovation and the fact that its storage protection and architecture prevent things like memory leak between individual users on the system at any one time. T he mainf rame continues to evolve, rea lly in ever y parameter. Just as processor s get faster in a lter native architectures, so do they in the mainframe. The difference being the design point in the mainframe is again one focused on sharing resources. So they have very large cache systems and high-performance memory systems, all designed to make the mainframe an outstanding resource to share across many users. Another parameter people associate with the mainframe, correctly so, is robustness. “It just doesn’t fail.” That’s a statement in the hardware, the OS and the middleware. T he nice t hing about mainf rame is we’ve developed relationships wit h customers. T hat’s dr iven us f rom a requirement standpoint. We continue to evolve the mainframe to meet the most demanding requirements of our customers. Long-standing customers continue to drive us forward. New customers in emerging markets like India and China realize the benefits of mainframe and we’re beginning to evolve to meet their requirements too. Are there any mainframe uses, whether customer driven or not, that have surprised you? People seem to think of the mainframe as a static entity that hasn’t changed in a long time. Nothing could be more incorrect. The mainframe continues to evolve ver y substantially and has since its inception 45 years ago. People see the industry moving ahead yet don’t understand the level of innovation that’s occurred in the mainframe. Talking about virtualization, the mainframe invented virtualization and holds many of the patents involved in the concepts of virtualization: the notion of a virtual machine, the notion of having a completely shared I/O set of resources. Those things indicate the kind of innovation that occurs in the mainframe. If you use Windows* operating system and many times see a blue screen, many of those problems are memory leaks that the intrinsic System z architecture with storage M A R C H/A PRIL 2009 Q: A: One thing that’s interesting, and I don’t shock easily, is work we’ve been doing with the Brazil-based gaming company Hoplon. Using the mainframe in the gaming world strikes me as odd. It’s an example of a future direction, which is adding unique acceleration capabilities around some very specific things we want to do. In the case of Hoplon, they want the security because the game software is proprietary and gamers try to break things. They want security, yet they want capability of high-performance computing. So they combine the capacity of the mainframe with the cell processor for numerically 26 ibmsystemsmag.com/mainframe

http://www.ibmsystemsmag.com/mainframe

IBM Systems Magazine, Mainframe - March/April 2009

Table of Contents for the Digital Edition of IBM Systems Magazine, Mainframe - March/April 2009