February 11, 2011
3000's Digital rival bids its founder farewell
Ken Olsen died this week at age 84. A certain generation of 3000 expert needs no introduction for Olsen, the man who founded Digital Equipment Corp. He was the last of the generation of business computer titan founders, preceded in death by IBM's Thomas Watson Jr., Dr. An Wang, Bill Hewlett and David Packard. Olsen's company took Hewlett-Packard and the 3000 to the mat in the 1980s, but eventually showed that even a brilliant leader of engineers can have blind spots fatal to a company.
HP 3000 vets were sharing some stories this week on the news that Olsen died, returning to the era when a minicomputer was still a known commodity in the IT enterprise -- what we called Data Processing back in those days. In the middle 1980s DEC had stolen the march on HP and its nascent PA-RISC designs, simply by having shipped VAX systems that already had the coveted 32 bits worth of chip bandwidth. Digital had trumped its beloved PDP systems (above) with a revamp that powered the VAXen. "Digital Has It Now," the ads boasted in tabloid newsweeklies, some printed on a silver ink background.
But Olsen's myopia matched his company's visions about clustering (still better than most competitors) and chip architecture (Alpha never deserved to be put down by HP once it acquired DEC). Olsen never said computers didn't belong in the home, but didn't figure them to be dictatorial controllers of the house needs like HAL in 2001 or worse. There are other comments to burden his legacy, like labeling Unix as "snake oil," to defend DEC's crown jewel of an OS, VMS. As it turned out, VMS earned its current day slot in the HP lineup -- Enterprise Business OS That's Not Snake Oil -- at the expense of the 3000. But the DEC lineup also yielded a product that funds the development of new 3000 hardware, even today, in an indirect way.HP looked at several factors when it decided to drop the 3000 from its future back in 2001. But not least among them was purchasing almost a half-million customers using VMS, which was in direct competition to MPE/iX and the 3000. Right as the [DEC-owner] Compaq announced it would be acquired by HP, analysts tool note of the overlap in the two business product lines. Olsen had led a company that built a worthy competitor to the 3000, sparked by customer zeal to match an MPE advocate's. In 1986, Fortune magazine named Olsen "America's most successful entrepreneur." DEC was 20 years younger than HP and established as a computer dynamo. It had climbed into leadership as the IBM alternative by embracing the downsized business computer: A kind of king of the mini, with HP bidding to overtake that throne.
During that year among those silver-papered days, DEC went straight to the carotid artery of the 3000 in a battle. The computer of the 3000's future was late -- MPE/XL was still crashing and slow, and everyone could see HP wasn't going to meet a 1986 ship date for the 900 Series. HP users went to that fall's Interex conference in Detroit with the hope that Hewlett-Packard could put their fears to rest. The Series 70 systems built upon CISC were not meeting top hardware demands.
HP didn't make many delivery specifics available at that show, while DEC invited users one and all to a special suite across from Cobo Hall. Up on the 11th floor, a handful of VAX systems were running what amounted to an IMAGE lookalike. DEC had even gone to the trouble of naming some datasets with strings like "rego," a nod to how important the Adager creator Alfredo had become to 3000 databases.
The Digital panache probably didn't sell many conversions -- that's what we called such a move in the '80s. But the ploy did spark a rapid response from the HP sales force attending the conference. Digital had called its conversion offer the Systems Attract Program. The day after the suite opened up, HP's salespeople were on the floor of Cobol Hall with badges: "Don't be a SAP for DEC."
It took five years or more, but eventually the RISC 3000s started to lop off some of the DEC business for enterprises. Up to the point of that face-off at Detroit, it was easy to say that the default for computing at a company was IBM or DEC, both larger than HP's efforts or any of the BUNCH companies (Burroughs, Univac, NCR, Control Data and Honeywell.). Further off the mainstream were Wang and Data General. All were alive and kicking in that fall when the 3000's future for the '90s was stalled.
DEC never tapped the PC windfall the way IBM did, missing opportunity even more badly than HP did with its TouchScreen PCs. It added a Unix solution late in that derby, so didn't have that nouveau technology ready to tout as "open." When the company had skidded on bloated product cycles it got itself bought up by Compaq, a tiger of an owner which devoured the Olsen's DEC culture as completely as Compaq overwhelmed the HP Way in the subsequent merger.
But the DEC heartland, VAX and PDP and VMS, have survived so thoroughly that one HP 3000 player has made enough money to fund an emulator project off DEC revenues. The European Migration Center for Digital bought itself out when Compaq purchased DEC, becoming Stromasys after a name change. Superior designs don't go out of style in IT shops where value trumps popularity. Even the PDP systems, which are in the background of many HP 3000 leader's resumes, are emulated today. Olsen can take complete credit for leading a company whose product was so popular it could spin a fresh future for a rival system: your HP 3000.
Use our search engine to find 20 years
of HP 3000 news and articles
The comments to this entry are closed.