The Longevity of the Mainframe

In this podcast Gregg Willhoit provides his ideas regarding the longevity of the mainframe. The podcast last for 4:40.

You can listen to the podcast here: http://blogs.datadirect.com/media/GreggWillhoit_LongevityOfMainframe.mp3

If you’ve been in our business – the mainframe business – for a while, you’ve read so many stories about mainframes coming and going. You know, the demise of the mainframe, and even more recently the good side of it, which is the resurgence of the mainframe. But it’s kind of interesting if you try to predict the longevity of the mainframe, or just to think about it, of course the mainframe platform – the IBM mainframe platform – has been around forever; for quite a long time anyway. MVS, MBT – all those variance, but basically the original operating system which ran on IBM 360 up until now with z/OS and the z10 series, what’s amazing to me is not only just how long this platform has lasted, but just how well IBM has treated their customers, especially with regard to downward compatibility. I’ve yet to see any platform that even comes close to allowing existing investment in IT be carried forward without any modifications for that matter to legacy programs. It’s just really amazing.

One of the things that I started thinking about when thinking about the longevity of the mainframe was, I think it was Thomas Watson Senior’s quote in the ‘40s where he predicted that “the world only needed 5 mainframes.” I’m not sure if that’s an exact quote, but I’m pretty sure it’s close. And what is interesting is how prescient that quote is. What I mean by prescient, obviously right now there’s more than five mainframes. But if you look at the direction that we’re going into with this whole grid based computing and cloud computing, maybe we’re really going to this only five computers in the world scenario. Maybe Thomas Watson is actually going to be right in some way. Maybe there will just be five huge clouds.

Now speaking of clouds, I think that the mainframe is extremely well positioned for a cloud computing environment. Any sort of cloud computing architecture scenario requires lots of the things that the mainframe is really strong on. Specifically virtualization – the ability of a platform to support multiple operating system images, each owned by possibly different companies or groups of companies running different operating systems and what now – virtualization is a big deal. And there is no platform that even comes close to the mainframe with regard to how efficiently virtualization is managed. And then, another thing with the cloud, you have to have extremely – I think you do anyway, maybe not so much in the nascent stages – but as computing platforms evolve from just the early adopters to the real significant users – that is when a significant portion of businesses use these platforms – you need management, dynamic workloads. The ability to manage work as it flows throughout the cloud, if you will. I think there is only one vendor that’s really poised and has the architecture to handle that, and that’s IBM as well with WLM and WLM based management tools.

I think the longevity of the mainframe at this point, I think it’s going to last longer than any platform that is out there. Obviously it has so far. I mean look at the mainframe. Look at the old days when there was a mainframe, it was very centralized, and then dumb terminals were on top of everybody’s desktop or people that had access to the mainframe desktop. There was real software that resided on these desktop appliances. Everything was done on the mainframe. I actually think that we’re going to go back to that kind of scenario where virtual desktops are probably loaded dynamically from these large mainframe servers onto a network appliance, so that these virtual desktops appear almost instantaneously. They eliminate the need for us, as computing power users, to manage all the various applications, worry about the hardware upgrades and whatnot. All the issues that exist today with standard PC based software.

I think that the cloud, virtualization and all that entails, portends a really great future for the longevity of the mainframe. One of the aspects of longevity is the ability to adapt. Sort of like Darwinian Theory. And the mainframe has showed a remarkable ability to adapt and survive. I think that specialty engines are an excellent example of that ability. I believe specialty engines have had, and will continue to have in an increasing fashion a significant effect on the longevity of the mainframe.