> Well, 5,000 would impressive in that it would be a lot more than some > people realize. <> I mentioned that a technology VC said to me > "mainframes? Does anyone still use those? Does anyone still make them?"
Yeah, there's no shortage of ignorant dolts out there. > I'm not sure I'd want to know. Every sales and marketing organization of every ISV would sell their first born children to get access to that data. Nobody outside of IBM has it and as someone else noted, IBM is notoriously coy about it. > If it came down to say every one of the Fortune 5,000 (although some > might have multiple sites) The largest customers each have many individual machines and even more OS licenses. The top twenty or so customers might even account for a majority of MIPS shipments. The top hundred combined might also account for the majority of machines and licenses. Maybe, I don't know but I can speculate based on our own customers. There's a fairly steep slope from the high to the low end. There are a lot of customers on small systems, hence the demand for capacity granularity on the z9 BC machines. The thing nobody outside IBM knows for sure is how many there are and who are they. The number of machines in existence is probably declining (I suspect) but they are on average getting bigger. Makes sense when you realize a single current z9 engine has more than ten times the horsepower of any (IBM) ECL engine and most of the first few generations of CMOS. IBM's ability to deliver compute power has rapidly outstripped all but the largest customer's ability to consume it. Absent some significant new workload growth drivers, they are going to massively overshoot. So it is no surprise to see all of these initiatives for new workloads on the box and the funny games with new engines and new pricing models. > would you really consider it a selling point > that there were only 5,000 shops left (or even 10,000) vs. the > googolplex of wintel/*nix shops? I guess it depends on what you're looking for. If you are a decision maker and you care about business value, then the sheer volume of monetary exchange that rides on mainframes might give you pause. OTOH if you are just looking for a cheap and easy way to put up a web site there's almost zero chance that a z9 of any size is on your shopping list. Could you do it on a mainframe? Well sure, but the market place has overwhelming voted on that already. With respect to differentiation based on the Fortune x000 (pick any first digit you like) they all look the same. They almost all have a mix of everything, so the fact that the ten largest businesses in the world are mainframe customers cuts less ice because they are also customers of Windows, Linux, *IX etc. Ask yourself if your own company could get by without all the other parts of their IT stack. Chances are they couldn't. That's why surveys that argue about the cost per user are basically flawed. CC ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
