Data General
|
|
Data General was a pioneering firm in the minicomputer world, known primarily for their Nova 16-bit minicomputer which gained a wide following and was used in many forms for over a decade. A series of missteps and bad releases almost killed the company in the 1980s, but they continued as a shell of their former greatness into the 1990s, eventually being bought out in 1999.
| Contents |
History
Origin, founding, and early years: The Nova and SuperNova
Data General (DG) formed when, in traditional computing style, several engineers from local companies got fed up with management and left to form their own company. In this case the main protagonists were Edson De Castro and Henry Burkardt III of Digital Equipment Corporation (DEC), and Herbert Richman of Fairchild Semiconductor.
De Castro was the chief engineer on the PDP-8, DEC's line of inexpensive computers that created the minicomputer market. It was designed specifically to be used in lab equipment settings, and as the technology improved was able to be shrunk to fit into a 19-inch rack, where many still operate today, decades later. de Castro was convinced he could do one better, and started work on his new 16-bit design.
The result was released in 1969 as the Nova. Like the later PDP-8 machines, it was a rackmount machine, but smaller in height and considerably faster. Launched as "the best small computer in the world", the Nova quickly gained a huge following and made the company flush with cash.
The original Nova was then quickly followed by the faster SuperNova, and then by a slew of minor versions based on the SuperNova core. The last major version, the Nova 4, was released in 1978. During this period the Nova generated 20% annual growth rates for the company, which became a star in the business community, and generated $100 million in sales in 1975.
Late 70s to late 80s: Crisis and a short term solution
The Nova had been supplanted in 1974 by their upscale 16-bit machine, the Eclipse. It was based on many of the same concepts as the Nova, but included support for virtual memory and multitasking more suitable to the small office than the lab. It was also packaged differently for this reason, in a floor-standing case the size of a small fridge. Production problems with the Eclipse led to a rash of lawsuits in the late 1970s, after new versions of the machine were pre-ordered by many DG customers, and then never arrived. After over a year of waiting many decided to sue the company, while others simply cancelled their orders and went elsewhere. It appeared that the Eclipse was originally intended to replace the Nova outright, and the Nova 3 series released at the same time was phased out the next year. However, strong continuing demand resulted in the Nova 4. The Nova 4 might also be a result of the continuing problems with the Eclipse.
In 1976 DEC announced the VAX series, their first 32-bit minicomputers, which they described as super-minis. The first products would not be released for a few years, but that would be just when current 16-bit machines would be getting old enough to replace. DG immediately launched their own 32-bit effort in 1976 to build the world's best 32-bit machine, known as the Fountainhead Project. However when the VAX 11/780 was released in 1978, Fountainhead was nowhere near ready to deliver a machine, largely due to problems in project management. DG's customers quickly started leaving for the VAX world.
Data General then launched a crash 32-bit effort based on the Eclipse, known as the Eagle Project. By late 1979 it became clear that Eagle would deliver before Fountainhead, and an intense turf war started inside the company for the ever-shrinking budgets. Meanwhile customers were abandoning DG in droves, driven by both the delivery problems with the original Eclipse and the power of the new VAX. The project was written on in depth in Tracy Kidder's Pulitzer prize-winning book, The Soul of a New Machine (see references), making the MV line the best documented computer project in history, at least from a human perspective. In two short years the first results of the project were released in 1980, the Eclipse MV/8000.
The MV systems generated an almost miraculous turnaround for Data General. Through the early 1980s sales picked up, and by 1984 the company was over a billion dollars in annual sales. Yet in fact the MV was too little, too late. At the same time the microcomputer was rapidly making inroads to the lower-end market, and the introduction of the first workstations wiped out all 16-bit lab machines, once DG's best customer segment. While the MV series did stop the erosion of DG's customer base, this now smaller base was no longer large enough to allow DG to develop their next generation when they needed to.
Data General-One
Data General's introduction of the Data General-One in 1984 is an interesting side-note, one of the few cases of a minicomputer company introducing a truly breakthrough PC product. To put it into context: The 1983 Radio Shack TRS-80 Model 100 was a truly battery-operated, portable, and operable computer resting in one's lap—but had an 8-line text-only screen, a proprietary OS, and no floppy. IBM's 1984 Portable PC was comparable in capability with desktops. It was not a laptop, however, but an AC-powered luggable like the earlier Compaq (Compaq would not introduce a true laptop until 1988).
In contrast, the nine-pound battery-powered 1984 Data General One ran MS-DOS, had dual 3½" diskettes, 79-key full-stroke keyboard, 128K to 512K of RAM, and a monochrome LCD screen capable of either the standard 80×25 characters or full CGA graphics (640×200). In other words, it was a true laptop that was truly comparable in capabilities to desktops of the era. For the first time, it was actually possible to travel with a battery-powered portable that provided a software environment similar to the one on a typical desktop.
The DG-1 was, however, only a modest success. One problem was the use of 3½" diskettes, which were slightly ahead of their time; popular software titles were not available in 3½" format and this was a serious issue because then-common diskette copy-protection schemes made it difficult to for users to copy the software into that format. Although Creative Computing termed the price of US$2895 "competitive," it was a very expensive system and usually-needed additions such as more RAM and an external 5¼" drive drove the price higher yet. But the Achilles heel was the LCD display itself, which was not backlit, had low contrast, and was frequently accused of serving better as a mirror than as a screen.
Lock-in or no lock-in?
In 1988 two company directors put together a report showing that if the company was to continue existing in the future, DG would have to either invest heavily in software to compete with new applications being delivered by IBM and DEC on their machines, or alternately exit the proprietary hardware business entirely. Proprietary hardware, and its associated "lock in", provided DG with a customer base who were forced to come back to them for other products. Customers hated lock-in, but all the companies did it so there wasn't much that could be done about it.
In the past, most "big iron" machines ran custom software generated on-site by development teams working for the end customer. Lock-in wasn't a huge concern; the best machine that could run the task was often the only thing to consider. However, the introduction of SQL databases changed things considerably, the programming teams were no longer developing everything from scratch, but instead taking a number of existing packages and wiring them together. The decision to choose one machine over another changed, now the customers were looking for the platform that ran all of the tools they were going to use -- hardware was becoming a secondary consideration to software.
At the time, IBM and DEC both had major software development efforts underway, delivering their own SQL servers (DB2 and DEC SQL respectively), as well as various message queue systems that acted as the wiring within a program. DG simply didn't have the same sort of software development infrastructure as DEC or IBM, and was not prepared to fight this war without serious investment.
Thomas West's report outlined these changes, and suggested that the customer was going to win the fight over lock-in. They also outlined a different solution; instead of trying to compete against the much larger IBM and DEC, they suggested that since the user no longer cared about the hardware as much as software, DG could instead deliver the best "commodity" machines instead.
Specifically, DG should instead examine the Unix market, where all of the needed software already existed, and see if they could provide compelling Unix solutions. Now the customer could run any software they wished as long as it ran on Unix, and by the early 1990s, everything did. As long as DG's machines outperformed the competition, their customers would return because they liked the machines, not because they were forced to – lock-in was over.
AViiON
de Castro agreed with the report, and the future of the MV line was killed off. In its place, DG released a technically interesting series of Unix servers known as the AViiON line. In order to keep the prices low, the systems used the Motorola 88000 RISC chips, a commodity processor with fairly high performance. However, anyone could build a 88000 based machine, and its performance was not up to par with existing minicomputers. To deal with this problem, the AViiON machines supported NUMA-based multiprocessing, allowing the machines to scale upwards in performance by adding additional processors.
Another key element to all larger computer installations is high speed storage. At the time, commodity hard drives could not offer the sort of performance needed, but DG solved this problem in the same fashion: by running a large number of drives in parallel, the overall performance could be greatly improved. The result was the CLARiiON line of storage systems, which also offered much better price/performance than competing solutons.
Used together, the AViiON/CLARiiON combination delivered systems that outperformed traditional minicomputers of the same generation, something many in the industry thought was not going to happen so soon.
The final downturn and the takeover by EMC; life after death
After betting the AViiON line on the Motorola 88000 they found that the true commodity processor, the Intel 8088, was quickly running away from them in performance in the early 1990s. They then moved to Intel based machines, but the lack of lock-in haunted the company, and customers were able to choose from a number of similar platforms.
But CLARiiON did better after finding a large niche for Unix storage systems, and its sales were still strong enough in 1999 to make DG a takeover target. EMC Corporation, a major data storage company, bought DG and killed off everything but CLARiiON, which remains a major product line to this day. On the World Wide Web, all that remains of Data General are some EMC web pages at DG's old web address ( http://www.dg.com ), which only mention the old company's name in passing.
Software
Data General wrote operating systems for its hardware: RDOS for the Nova, AOS/VS and AOS/VS II for the Eclipse MV line, and a modified version of System V Unix called DG-UX for the Eclipse MV and AViiON machines.
They also produced a word processor, CEO, a relational database, INFOS II, and a programming language, Data General Business Basic.
Notable alumni
- Jean-Louis Gassée was with Data General in France before moving to Apple Computer and Be Inc..
- Edward Zander was product marketing manager at Data General before his career at Sun Microsystems and Motorola
- Other alumni are listed in Soul of a New Machine, below.
References
- Kidder, Tracy (1981). The Soul of a New Machine. Little, Brown and Company. Reprint edition July 1997 by Modern Library. ISBN 0679602615.
External links
- Data General's former Web site (now EMC's AViiON page) (http://www.dg.com)
- SimuLogics ("dedicated to preserving the history and legacy of the Data General Nova, Eclipse, MV and compatible computers") (http://www.simulogics.com/)
- Carl Friend's Computer Museum (has pages for over a dozen DG systems) (http://users.rcn.com/crfriend/museum/index.shtml)
- Guide to DG's AOS/VS System, part 1 (http://members.chello.se/brickhouse/vsthings/guide2aosvs1.txt)
- Guide to DG's AOS/VS System, part 2 (http://members.chello.se/brickhouse/vsthings/guide2aosvs2.txt)