IBM: System Z Is Not A Cash Cow
Other systems can’t match mainframe reliability, says IBM’s Jim Porell, and he dismisses any talk of IBM abusing a monopoly
When HP likened IBM to a heroin dealer on this site, we were not at all surprised to get a polite phone call from Big Blue and, within a few days, a distinguished engineer was explaining to us why mainframes are lot more healthy than Class A drugs.
Hewlett-Packard’s server launch last last month was wrapped up in a pitch for “simplification”, which turned out to mean switching off unecessary applications and – where possible – mainframes. Have us do it for you, said HP’s Paul Evans, because going to IBM would be like going to your drug dealer to get you off heroin.
System z … it’s good for you!
So, we found ourselves talking to IBM Distinguished Engineer Jim Porell, who works on System z Business Development. Among other things we touched on the fact that HP does its own dealing. Its services arm, EDS, already owns and manages many mainframes on behalf of customers like Aviva, and buys new ones from IBM to service those contracts.
He also dismissed the idea that IBM’s mainframe contracts are exploitative, and that it is abusing a monopoly.
But before we got there, we talked about the servers. Mainframes are not what they used to be, said Porell.
“Mainframes used to be blind and deaf. The human-computer interface is in things like mobiles, ATMs, kiosks and web browsers,” he said. “One of my major roles is how we seamlessly connect to those environments, in addition to managing ourselves better.”
Quickly, he rattled off the areas where mainframes have a reputation for being “better”.
- business resilience
- workload capacity
- business process integration
- system management
As well as that reputation, he admitted, mainframes had an image of being old and boring which IBM – like others such as CA along with BMC and Compuware – is doing its best to dispel: “We have done a number of things to open up the interface: Linux on the mainframe, C++, Java, Enterprise Java Beans. We can take the same code on many distributed platforms and run it on the mainframe.”
“We’ve spent a couple of million over the last few years reconstructing the interface, and increasing the level of skills portability,” he said. “If you are used to point and click, those skills should be portable to a mainframe.” The new-look for mainframes is also coming from IBM’s Tivoli management outfit.
