Intellivision III

MASTER COMPONENT [UNRELEASED]

DEVELOPMENT HISTORY

When Intellivision was introduced, Mattel's marketing positioned it as the cornerstone of a home computer system, implying that while the Atari VCS (2600) was a toy, the Intellivision -- Intelligent Television -- was an educational tool. After its release though, Marketing discovered that most people who chose Intellivision over Atari did so because of better graphics. Educational games and the Keyboard Component were dropped in priority and the focus shifted to exploiting the graphics with a series of commercials showing side-by-side comparisons of Intellivision and Atari games.

This backfired when Colecovision was introduced in early 1982. Suddenly, Intellivision was no longer the superior graphics system. Atari tried to compete with the introduction of the Atari 5200, but consumers were disappointed it couldn't play Atari 2600 cartridges.

Marketing decided that the only way for Mattel to respond was to introduce a new game machine -- Intellivision III -- that had graphics as good or better than Colecovision and could still play all the original Intellivision cartridges.

The solution to this was actually fairly simple since the Intellivision contains a separate video processor, the STIC chip. Mattel commissioned General Instruments to build an improved STIC for the Intellivision III. This new Super-STIC (STIC 1B) would have a double-resolution background mode and allow for more moving objects (sprites) and colors, but otherwise would be compatible with the original STIC. By having the Intellivision III based on the same CP1610 processor as in the original Intellivision, the old games would still run, and new games could take advantage of the improved graphics features.

General Instruments prototyped the Super-STIC, Design & Development modified an Intellivision to use it, and APh Technology Consultants began writing the expanded EXEC program to control it. The project was code named Coffee.

Had that been the extent of it, an Intellivision with a new STIC chip, expanded EXEC ROM and some extra RAM (to keep track of all those moving objects) -- the product probably could have hit the market fairly fast. Unfortunately, Intellivision III fell victim to Kitchen Sink Syndrome, as in Everything-But-The.

While Intellivision III was in development, Intellivoice hit the market. Consumers liked the concept of talking games, but they didn't like that they had to buy an add-on module. Sales were sluggish. To save the voice program (which represented quite an investment), it was decided to incorporate Intellivoice into the Intellivision III.

Now, the Intellivoice contains a buffer chip to interface the speech processor with the Intellivision CP1610; it was designed so that other peripherals could also interface with the CP1610 through this buffer chip. On the drawing board at the time: wireless hand controllers. Since the buffer chip would have to be built into the Intellivision III, it was decided to add the (as yet undeveloped) wireless hand controllers, too.

Of course, if you're going to double the graphic resolution, you really should double the audio quality, so an extra sound chip was added to the design. And, since the input ports of the Intellivision sound chip are used for the hand controller inputs, that meant you could add two more hand controllers to the Intellivision III and design 4-player games.

All of these extra features meant the new EXEC had to be even more complex to control everything. Not only that, but the Intellivision programmers wanted to see frequently-used subroutines, such as screen scrolling, added to the EXEC instead of having them use up precious game cartridge space. Ray Kaestner (Burgertime) was sent to APh to represent the Mattel Electronics programmers during the development of the new EXEC.

These expanded design features were grafted on one at a time over several months, causing a lot of rework and frustration and delays.

In private rooms in the Mattel Electronics booth at the June 1982 Consumer Electronics Show (CES) in Chicago, major toy buyers were told of the upcoming improved Intellivision to bolster their faith in the product line. At the January CES in Las Vegas, they were told, they would see the working system. But by January 1983, Intellivision III still hadn't progressed beyond the preliminary breadboard stage in the Design & Development lab.

So if it never left the lab, what were the toy buyers looking at in the private rooms of Mattel's booth at the January CES? Not a prototype Intellivision III as they thought. They were looking at a plain old Intellivision displaying some really good graphics.

Six cartridges were shown; two were supposedly games in progress, the other four demonstrated enhanced features of Intellivision III. The nearest thing to a real technical advancement in these cartridges was that they contained up to 16K of memory. Since they were all graphics, special effects and music (by Bill Goodrich) and no game play, they could be a lot flashier than the then common 4K real game cartridges.

The two "games in progress," shown with printed packaging, were Treasure of the Yucatan and Grid Shock. The first was a static picture of a stone idol overgrown with jungle vines. An impressive, complex screen, it had been done a year earlier by Eric Wels (Mr. Color) when he was first hired, simply to learn how Intellivision graphics worked. The screen eventually found it way into Bill Goodrich's D&D; voice game, Quest.

Grid Shock was the beginnings of an actual game by Andy Sells. A wall that swept back and forth across the playing field, changing perspective as it moved,gave the screen a strong 3-D feel. Grid Shock had been abandoned by Andy since he was spending so much time doing sound effects and music for other games (e.g. Shark! Shark! and TRON Solar Sailer), but what was complete was visually interesting enough to pass as Intellivision III.

The other cartridges, written by Ray Kaestner and programmers at APh, used sleight-of-hand to demonstrate Intellivision III features -- multiplexing moving objects put more than the normal limit of eight on screen at one time (albeit flickering); updating moving object positions every 1/60 of a second instead of the EXEC's normal 1/20 gave the illusion of smoother, faster motion.

So what games were really in development for the Intellivision III? Well, none. Since both systems were CP1610-based, it was decided to just keep writing for the Intellivision. Then, when (and if) the features and release date of Intellivision III were finalized, any Intellivision games nearing completion would be quickly upgraded for the new system by tossing in fancy graphics and sounds. In October 1982 for a meeting with Marketing and distributors, Gabriel Baum, VP of Application Software, listed the likely candidates to be released as Intellivision III games: the then-in-development Basketball II, Mission XThin Ice, Air Battle and Mystic Castle; proposed Winter Olympics and Dungeons & Dragons cartridges; and a to-be-determined children's title using one of the newly acquired licensed characters.

But no upgrading was ever needed, because in mid-1983 Intellivision III was killed, done in by the delays. Retailers saw it as too little, too late to compete with the then year-old Colecovision. And with the Aquarius Home Computer System and the Intellivision Entertainment Computer System (ECS), there was already a glut of hardware in the 1983 pipeline. With Mattel Electronics starting to pile up hundreds of millions of dollars in losses, it was announced that Intellivision III was being canceled, officially because most of its features had been incorporated into the ECS. (A bogus claim; the extra sound chip and hand-controller ports are the only features they share.) The last hope for the future of Intellivision now rested with the top-secret project code-named Decade: Intellivision IV.

NOTE: The Intellivision III should not be confused with the INTV System III, which was simply INTV Corporation's re-release of the original Intellivision Master Component with minor cosmetic differences. To make it worse: at the January 1987 CES, INTV Corp. announced the INTV System IV, which shouldn't be confused with the Intellivision IV. The INTV System IV was really the Intellivision III; Glenn Hightower of APh had convinced INTV's Terry Valeski that the system was still viable. INTV Corporation, however, was not; despite making the announcement, they didn't have the financial resources needed to actually resume the development that had stopped three-and-a-half years earlier.