Bob Labant, IBM director of Application Business Systems and responsible for AS/400 development, was in the UK last week. He arrived hot on the heels of his marketing director, Mark Ryan, who was here in mid-September to tout the benefits of the new low-end machines, and to convince the sceptics that AS/400 users are the happiest bunch imaginable (CI No 1,514). Perhaps he didn’t do a very good job since Labant found it necessary to reiterate most of the facts and figures that prove the AS/400 is irrefutably successful. The rate of installation outstrips that of System 36s; double digit revenue growth during 1989 and the first half of 1990; market share is increasing; the number of new installations is on the up; the high-end machines are ranked second only to the 3090-600 in the US; and since it’s IBMspeak, customer satisfaction is high. Between 40% of installations are migrations from Systems 34, 36, and 38, 30% is from multiple installations, and 30% is new business. Of that last figure, half were non-IBM houses, and the rest didn’t have a computer system. In terms of worldwide revenue, 45% is derived from Europe, and Japan is the fastest growing market, although that is partly explained by the negligible base. Labant is perhaps more willing than his subordinate to acknowledge that the history of the AS/400 has been a chequered one, and IBM has not always displayed whole-hearted commitment. For example, he says that Systems Application Architecture compliance was neglected and that more should have been done more by now. To remedy that, he has reorganised the Rochester, Minnesota development site and appointed Dave Schliker – the OS/400 architect – as head, alongside Burl Wheeler, Mr SAA. Labant believes that IBM needs to get more SAA-compliance into the AS/400, and that it is now committed to doing so. However, he is more ambiguous on AD/Cycle. While the Synon alliance is a key application driver, Labant says that IBM is working on how much of AD/Cycle should be in the AS/400, and the issue is whether or not users really want an application-driven machine. IBM’s neglect of System 36 users has been a sore point which the company tends to deny, saying that it cannot understand such accusations. Nonetheless, when asked if he would be exerting himself to persuade these users to ugrade, Labant’s response was an unequivocal darn right, and he also admitted that – yet again – IBM should have done more before now. System 36 emulation ought to have been better, and the AS/400 should be as easy to operate as System 36s. He reckons that both of those problems have been addressed, and that more users will migrate as their businesses will demand and benefit from AS/400 applications. Also important is the openness of machine level instructions which provide the capability to customise applications. Changes in the mid-range arena, and user demands for an enterprise-wide system means that IBM will have to keep expanding the AS/400 in terms of power, communications and openness. Labant says that the AS/400 will cover a range of requirements, and there is no technological wall at the high-end. It is a lorry to an 18-wheeler system. Next year’s June announcement will double performance, and plans are afoot to go well beyond that in the following 18 months. All of which sounds as though the AS/400 could be in serious competition with some of the ES/9000 mainframes. Labant acknowledges that there will be an overlap – there already is with the 4381s – and he says that the AS/400 won’t be in that marketplace to cause conflict, but because users are demanding more power without the expense of maintaining data processing departments. Around 40% of AS/400 sales are to businesses with between 50 and 1,000 employees, 25% go into small businesses, and corporations with over 1,000 staff account for the remaining 35%. Since it is this last category that may well have a 3090 machine or equivalent, it seems that the AS/400 already has a foothold in the large systems marketplace. Interoperability, never an IBM favourite, has acquired such a head of st
eam that TCP/IP, Ethernet and a more open Machine Level Instruction Set have had to become part of the AS/400 world. But Lavant believes that while some vendors talk of openness only in Unix terms, his users want more. Small businesses look for interoperability with PC-DOS, intermediate organisations want personal computer support plus multi-vendor interoperability, and the largest clients want personal computer support, multi-vendor support and System 370 interoperability. He denies that the C compiler for the AS/400 is a first step to doing a version of AIX Unix for the machine, and maintains that applications will continue to differentiate between proprietary and open systems.
RS/6000 confusion
The RS/6000 and AS/400 are creating some confusion, but he says that the RS/6000 has not impacted on AS/400 sales, simply because there are many things that the Unix machine can’t do. It is not a question of Unix capability, it just that users have to buy a database, systems management and network management facilities. Labant says that these are already part of OS/400 and require neither the installation of multiple systems and footprints, nor a dedicated management team. – Janice McGinn
