Every November ends with a special holiday, Thanksgiving. This important day commemorates the victory of shrewd and intolerant Pilgrims over gullible and charitable Indians. Many Pilgrim traditions – witch hunts, turkey killings, funny hats, public dunkings and a shortage of good parties in Boston – are still with us. But there is far more to November than Thanksgiving. It is also election time. Across this great nation, hundreds, nay thousands of users elect to add computing capacity during the fourth quarter of each year. They choose to install new machines or old ones, to buy or to lease, to favour one vendor or another. Like Thanksgiving, election time gives the Indians a rare opportunity to palaver with the chiefs. Unlike those at Thanksgiving dinner, the turkeys brought to the table by computer vendors are hard to identify until much later, when they are sold or subleased. At that time, not only the turkeys but also their owners may be killed. This year, more than for several past, there has been a lot of activity at the chopping block. Hiperspace babble The market’s treatment of new as well as used equipment promises even more of the same in coming seasons. It’s no wonder, then, that some prominent user organisations seem to have chosen election time for a vacation. For instance, one prominent insurer has told lessors that it elects to defer new equipment purchases until 1989. IBM and some third parties are still in there pitching, of course, and might yet loosen this user’s grip on its purse strings. But if this company’s view of the economy and its own corporate prospects is correct, holding off on big buys until next season could become a trend. IBM’s near term prospects are further dimmed by the bandwagon effect. The more that users believe the computer market has gone soft, the greater is the likelihood they will further curtail orders. This can be a wise user strategy. In the past, when IBM has gotten edgy about its fourth-quarter prospects, it has cut prices, rewarding the hesitant. Users’ present apathy toward election time shows that there is something wrong with IBM’s prime candidates, mainframes and disks. In response to the manufacturer’s impassioned campaign for its latest variation on the 370 architecture platform, a good many executives on the user side have done the very worst thing they could do to IBM’s earnest sales reps and faithful shareholders. They have ordered upgrades and new systems… but for easily-deferred delivery in the future. And in response to all IBM’s ESA, PR/SM and hiperspace babble, they also have said two words – to their colleagues, if not directly to IBM – that are rarely heard in the puritanical halls of corporate America: So what? The catch with IBM’s mainframes is not the required migration to yet another operating system. By most accounts, the move from XA to ESA is as easy as anyone could want. The issue faced by IBM’s most important customers is the cost of migrating to a 3090E or 3090S. Upgrades at $60,000 or so per MIPS only seem cheap if the investment lasts. When the resale value of that MIPS falls to half its cost in a year or two, and to a quarter of that amount 12 months later, the upgrade is mighty expensive. Most lessors, on the assumption that the upgraded 3090s won’t be worth much in the future, want users to sign full-payout leases on mainframe enhancements. This makes the annual cost of extra MIPS very high, and it discourages users’ tendency to speculate on additional computing power. This is always the case when a mainframe series enters the twilight years of its economic viability as a new product, and no mid-life kickers can alter that. The other big factor shaking users’ confidence is the age and fragility of the 3380 disk family. None other than Jim Porter whose Disk/Trend Report is the most respected reference in the disk industry, says in his October edition that IBM is likely to announce two disk families in 1989. Although guru Porter doesn’t philosophise about the used disk market from his Mountain View ashram, it seems clear to us that these ma
chines could send the 3380s, including the Ks, right to the boneyard. During the first half of next year, Porter predicts, IBM will announce a very fast 10.25 unit suitable for swapping and other demanding roles. The second half of 1989 should witness IBM’s foray into the world of grouped disks with 5.25 platters which it will gang up to provide large capacities. Both these products will be aimed at large mainframe shops, and both will depend on 3990-3 disk controllers for optimum performance. IBM’s problems with software for its 3990-3 disk controller will amplify any economic shocks the used disk market must absorb during 1989. IBM will install hundreds of partially functional 3990-3 controllers between now and its release of the control code it has promised users. Once IBM signs off on the code, it will be available overnight to the entire installed base of 3990-3s, instantly giving users compelling reasons to migrate to next generation disks, which will be storage devices designed for the 3990-3 and ESA-era software. The 3380s, including the K drives, are the best disks IBM could build for its XA generation. In other words, they, like the Pilgrims, are history. Even if IBM prices its new disks high at first to skim all the cream off the market, used 3380s will take a terrific beating. The big shops with the generous discounts will install the new disks as fast as IBM can ship them, displacing hundreds of 3380s. The disks displaced first won’t all be Ks; users will dump standards, Ds and Es first. But there will be more used Ks than investors in IBM’s most capacious 14 disks would wish for. Even now, aware that 3380-Ks are a risky if not suicidal investment, quite a few large shops have elected to slow or halt disk acquisitions. This has not yet dawned on the organisations that fund disk portfolios for lessors. Blue-haired grandmother Hundreds of millions of dollars in large disks have been purchased from disk lessors by income funds. These financial entities amass the hard-earned capital of trusting widows and orphans. They direct this flow of funds to such stalwart institutions as the used disk market and the second-hand terminal business. The investors are provided with voluminous cautionary paperwork that is only scrutinised after somebody takes a hit. Devoid of legalese, an income fund prospectus says that a deal is like a nose: sometimes it runs, sometimes it blows. Lessees can rejoice that some risks they or their lessors might otherwise have taken belong to purchasers of income funds’ securities. But some computer shops might have to tolerate visits by investors in unit trusts, which are vehicles that tie funds to specific pieces of kit. Reportedly, one industrial site has already been raided by the legal owner of a 3380 who wanted to confirm the existence and serial number of a disk drive. Few data processing managers would relish the sight of a blue-haired grandmother being carried out of a computer room clinging to a 3380 and crying her eyes out. That’s why some users have stopped electing to buy ageing machines at new iron prices, while others vote only for lessors that don’t sell their lease obligations to income funds. (C) Computer & Communications Buyer
Content from our partners
