Classic Computer Magazine Archive CREATIVE COMPUTING VOL. 11, NO. 11 / NOVEMBER 1985 / PAGE 86

Trade shows and other mysteries. (IBM images) Will Fastie.

About ten years ago, I went to my first computer trade show. I had to sweet-talk my boss; the company couldn't see a good reason for me to go and, at the time, neither could I. But go I did, and wow! What a time! The mother lode! I was in hacker heaven (I think I may have been a hacker back then).

Nowadays, when a trade show rolls around, I feel the way I used to when my mother yelled at me for the umpteenth time to take out the trash. Much of the glamour is gone, but I must go for very sound business reasons. This year I went to the National Computer Conference (NCC) and SIGGraph '85, the conference and exposition of the ACM's special interest group on graphics. The NCC was good, and SIGGraph was spectacular.


In the past, NCC was my show. Working, as I did, for a company doing systems integration, the need to understand the marketplace for commodity items (computers, disks, controllers, peripherals, etc.) was very great. NCC has always been perceived as a show for DP and MIS; I viewed it as a show for OEMs.

As the microcomputer revolution struck, I noticed that the conference proceeded apace, but the exposition became more DP-oriented. Two years ago I think NCC reached its zenith in that regard--there was very little interest in the OEM side of the market. Last year the show was an incredible bore, but this year, notwithstanding the conclusions of the trade press that the show was off, I thought it was returning to its roots. If I had been in my former life, I would have found much of interest. In this life, there at least were signs of future activity. Here is a quick glimpse.

* Laser Printers: There is some action here that may result in lower cost printers by the time fall COMDEX rolls around. Kyocera displayed a print engine that had a separate drum and toner cartridge. Toner is good for about 3000 copies, but the drum can last for about 10,000, lowering the cost of operation considerably. In the Canon engine (Apple, Hewlett-Packard), the drum and toner get replaced at about 3000 copies.

* Other Printers: Because NCC does not serve the retail market (that's COMDEX territory), only a few printers were on exhibit. However, I was surprised to see few new entries in the dot matrix and fully-formed character race, so again look for a busy COMDEX. It's hard for me to believe that companies still enter this crowded market dominated by Epson and Okidata.

* Hard Disks: Capacity and performance are up, and price is holding well. I paid the most attention to 5-1/4" drives in the performance class required by the PCAT, for which IBM specifies better than 40 milliseconds for an average access. Priam was the most interesting; after three years of work they have finally solved their linear motor problems and produced a high-performance device (30ms) in the 5-1/4" form/factor. Previously their drive had been too deep.

* 80286-based Machines: The AT has clearly sparked the market. Whether the machines are AT clones or not, IBM's choice of the next Intel processor establishes it as an important de facto standard. Rumors have it that Intel is working on a special version of the 286 for IBM that includes the 80386 memory management technique, and the 386 itself is a natural successor for future systems. Everybody's 286 box runs at least 8MHz except, of course, for IBM's. Drat.

* Tape Systems: Tape backup is becoming a hot market right now as companies begin to pay attention to basic DP issues long avoided by users of micros. Data integrity and security are well established for mainframes; the word is filtering down. However, there are several competing standards and methodologies that confuse the end user. A new IBM product, the 3480 tape system, introduces a single-spool cartridge using 1/2" tape and with a capacity of 200Mb that is more compact than the traditional Scotch/3M 1/4" cartridges in wide use today. The IBM system is for mainframes, but the existence of the cartridge is the important thing: look for clever third-party vendors to develop low cost systems around that cartridge. This may be the most important development shown at the NCC.

*Backup: One software product did catch my eye. It is FastBack, a backup program for the PC family from Fifth Generation Systems. It is a significant product because of its very high speed of operation, and it is faster than any other backup program I have encountered. The program performs backup to diskettes, just like the IBM BACKUP command, but it has a very nice visual interface and clever features, such as the ability to alternate between two disk drives automatically. A large number of PCs have been upgraded not only with a hard disk but also with half-height floppies; FastBack allows such configurations to be fully exploited.

That's about it. And, of course, that's why so many reporters called the show slow and boring. But I think my list represents significant future developments that we should keep our collective eyes on.

SIGGraph '85

Nothing in the computer industry dazzles quite as much or captures public attention quite as effectively as computer graphics. Witness our slavish devotion to arcade games and to their subsequent in-home version, the Atari 2600. Witness also our disdain for them now, as technology offers more and more and these old(!) things look so primitive, so simple. And note the technical success of "The Last Starfighter," a sci-fi melodrama whose special effects used not a single scale model: everything was done by computer-generated graphics. Something to see, and you should.

SIGGraph is the premier graphics show in the industry. It is hard to describe the excitement I feel as I walk the aisles, observing one miracle after another. From a technical point of view, the graphics capability of even low cost systems is rising dramatically, and this fact was more evident at the show this year than ever before.

I looked for two things. First, I wanted to see how many IBM PCs were on display as the computational engine underlying a graphics system. Second, I wanted to see how far the vendors had come with resolution, color, and speed.

On the first count, I struck gold, but not exactly with PCs. The IBM PC AT was everywhere, doing everything. It was the predominant desktop computer at the show, and it was ubiquitous besides. I could not even begin to recount the firms with product offerings based on IBM equipment; if this is an area of interest, you may be sure that you will soon hear from these companies. It is interesting to note, however, that few firms had products specifically for the AT. Most were built with the 8-bit bus of the PC in mind and just demonstrated on the considerably faster AT. I should also note that most of these products should work properly in AT clothes like the Compaq 286 which, at 8MHz, are 33% faster than the AT.

I looked for software products designed to take advantage of the IBM Enhanced Graphics Adapter (EGA) and its 640x350, 16-color capability. I saw almost nothing; most of the excitement surrounded products of greater screen resolution and more color choices. Although I think that the EGA is quite good and will become an important, much-copied standard, it represents a middle ground of capability in the graphics market. That means the EGA will be used for business graphics and visual presentations, but lacks the power to be used effectively in demanding, serious graphics applications. IBM has chosen something that can drop in price and be popular, while the graphics market needs devices that stretch the limits of technology.

The EGA is thus a good benchmark against which to measure other products. First, the EGA delivers a resolution of 640x350. Although that resolution is close to popular sizings like 400x400, 512x512, and 600x400, higher resolutions of up to 1024x1024 are becoming feasible and affordable. Some such displays were shown at SIGGraph; expect the vendors to move rapidly here. Second, the EGA delivers a maximum of 16 colors chosen from a fixed palette of 64. Here IBM falters; the open market is demanding and getting 256 from 4096 or even more. A vast array of colors is an important feature, perhaps more important than the resolution. With many hues it is possible to shade objects much more effectively, giving the illusion of smoothness even in coarse resolution. As the resolution rises, of course, such shadings get smoother, but then the color advantage can be applied to ever smaller objects.

SIGGraph was a hotbed of such display adapters. They certainly were superior to the EGA in resolution and color; in performance the comparison is not fair. Because the cost of microprocessors and memory is forever dropping, the manufacturers of graphics adapters are increasingly turning to the solution of a complete graphics subsystem on a plug-in board. The IBM Professional Graphics Controller (PGC) is an example in this genre, including its own 8088 processor and a complete onboard library of graphics display primitives. Within months of IBM's announcement of the PGC, several companies produced clones with much greater performance, due largely to the choice of better processors (80188, 8086, 80186) and faster clock rates. The general market is building devices that are not only faster but include even niftier features. These were in great supply at SIGGraph; expect to see software to drive these boards emerge over the next year.

Another area of interest for me is hard-copy graphics output. I expected more than I saw. In particular, I guess the market must be waiting for HP's LaserJet II (I'm guessing that it will include full-page graphics) before it pounces on laser-generated output. There were very few laser printers in evidence--a surprise. On the other hand, the price of pen plotters continues to drop, so conventional drafting applications should filter downward into smaller and smaller firms with the expected benefit.

Both shows were fun for me, and both had their share of surprises. Next year's NCC is in Las Vegas in June (better than July, but not much) so the heat may mitigate my objectivity. SIGGraph I will follow anywhere. Both are worth the investment for their target audience, and I look forward to reporting on them again.