Martin Banks, Personal Computer World 04/82 - checked

April 1982

One of the minor inconveniences of writing for a monthly magazine like PCW is the timescale involved in producing the beast. The words you read here were written towards the end of January, yet the mechanics of magazine production dictate that they do not appear until now.

This sometimes makes writing a trifle difficult, for occasionally something happens (or more specifically, looks as though it's about to happen) that is worthy of comment or prognostication at an early stage. By the time the words appear, however, the situation that prompted the prognostication may have changed.

That could well be the case with this month's subject: the first appearance of news about a new machine from dear old Commodore (which has surely been around long enough to be promoted to Rear Admiral). The machine is called the Commodore 64 and it has been waved around at the Consumer Electronics Show in the USA. In Europe its first showing is expected at the Hannover Fair.

Now, the problem is that information on the Commodore 64 - at the time of writing this - is somewhat scant and is likely to stay that way until the European launch, an event approximately coincident with the publication of this issue.

But even that would not be a problem in most normal circumstances, for what information there was available could be ignored until after the event and an article could be prepared then. Indeed, in most normal circumstances that is exactly what would happen. The Commodore 64, however, sounds interesting, both from what the early information suggests the company is trying to achieve and from the viewpoint of the company's products, post Chuck Peddle.

As most reader will by now be aware, the daddy of the Commodore PET, Chuck Peddle, left the company over a year ago to form his own show. That show is Sirius, a company that launched a comprehensive and competitively priced machine at the back end of last year. At the time of that launch, Chuck was questioned about his view of Commodore now he was outside the company. After all, since the PET, the company had come up with little in the way of internally developed new products. Most of the PET add-ons came from the approved-products scheme and even the VIC had strong Japanese ancestry.

His replies were interesting, especially in the light of recent events. `Commodore management is primarily interested in shifting product,' Peddle said. He followed that with the prophecy that if the company chooses to stay with its predominantly low-end spectrum, it would continue to be around for a while.

Well, recent developments from the company would tend to indicate that it does indeed to stay at the low end of the marketplace. The more long-term developments that the company appears to be making, moreover, will tend to keep it very much in the business of high-volume sales - `shifting product' as Peddle puts it.

The Commodore 64, on the current sparse information, does not sound spectacular, except for two points. One (almost of necessity) is price. In the US it will retail for $595 while providing a performance comparable to an Apple II retailing at around $1650. There is nothing amazing about that, really, for semiconductor technology is always driving down the price of systems, while maintaining their capabilities, or even pushing them up.

The second point is, in its own way, much more interesting and the previously mentioned comparison with the Apple II becomes more relevant, for the Commodore 64 can run Apple software. OK, so the software will have to be entered via the keyboard rather than from disk but Commodore reckons the necessary modifications for that function could be provided at negligible cost.

Such 'negligible cost' means that it would be an easy step for Commodore to strike at the very heart of its great US rival, Apple, and produce a low-cost emulator of the Apple II that ran all the software available for the US market leader and that at a price that could seriously damage Apple financially. It is, for example, interesting to note how, since adopting Apple and others as the star stock market purchases of 1981, American stock analysts have rapidly become 'street-wise' enough to notice this possibility from within the Commodore announcement and bomb Apple's stock price by $2 in a single day.

Commodore apparently says that it has no plans to do that small thing, ever. Instead the company would appear to be after bigger fish. "Bigger' in this context might sound derogatory towards Apple and as such this is not intended. Rather, 'bigger' refers to the whole damned microcomputer business, which now includes some other minor inconsequential companies like IBM.

For Commodore would appear to have latched on to the obvious idea that once you have learned how to emulate another type of microcomputer system and found a way of achieving the emulation that is both simple and inexpensive, then why the hell should anyone stop at just emulating one type of system? Why not emulate the world - or at least all the relevant bits of it such as Apple, IBM, Tandy, Sharp, NEC or anything else you fancy?

In hardware terms this is now becoming a relatively easy task. Several people in this country have been looking seriously into the possibilities of doing it, and products in this area may well appear in the near future.

The key appears to be the use of gate arrays - that group of microelectronic devices that allows circuit designers to produce economically a whole range of sexy custom-design devices for whatever they want to achieve (this was covered by the 'Statement' of October last year). The arrays provide a very useful facility for the system designer who wishes to emulate another machine; for there are several ways of doing this trick, and many problems to be overcome. That facility is convenience.

The two main options facing the designer are: a) use bit-slice processors and microcoding to emulate the instruction set of the emulatee; and b) interface the main system processor with the same system processor as the emulatee. The first route is often the more elegant, and one followed by many of the minicomputer manufacturers to emulate their own products. Its major drawback is that microcoding instruction sets is an extremely complex black-art, and is not really feasible for emulating a range of different machines. The second route is better because the emulatee processor(s) can be mounted on a circuit board, plugged into the main board of the basic system and away you go. Well, in theory anyway, for the two beasts have to be interfaced, and that means a gate array.

There are many subsidiary problems to be overcome after this decision has been made, and the gate array approach has a part to play here as well. The main problem is where different systems - and especially different operating systems - usually expect to find I/O devices, of different types, at different locations. This means interfacing both at the software protocol level and at the hardware device level. Again, this is the type of application for which gate arrays might have been invented.

It would appear that this is the route that Commodore has selected to follow. The Model 64 would seem to be the basic system for a new range of modular emulators that will almost certainly cover such systems as the Apple II, Trash-80 and the IBM Thingie. It would also appear that, as might be expected, the cost of each emulator will be fairly low - probably around the $1000 mark. Each emulator will be a module that plugs into the back of the Model 64 (or a variant/development of that basic machine). No doubt that machine will also have the capability for considerable memory expansion, for though the Model 64 has 64 Kbytes of RAM, it also uses the new MOS technology 6509 processor chip which has on-board memory management that allows up to 256k to be addressed. The processor chip also has on-board RAM, suitable for running a small set of commonly used instructions at high speed. This would be an ideal facility to have for controlling the operation of an add-on emulator board.

As one US stock analyst has already observed, if Commodore actually goes through with what it appears (and is rumoured) to be planning, it could dramatically change the shape of the microcomputer business, for it would have at last brought the hardware side of it down to the level of becoming an irrelevance. Potential users would no longer have to suffer the agonies of choice between rival and incompatible systems - a choice that immediately governs what software can be used.

Microcomputers, or more especially the variety of system emulators, would become akin to commodity items like cans of beans or Intel 8080 processor chips.

Whether this is a good thing for the business in the long term is open to question. The implication of such a move for the hardware companies is that they would all have to follow suit if Commodore's efforts proved successful. Prices would be savaged in an orgy of competition until such time as every potential user in the world was no longer potential - they were actual.

Now if such systems are as reliable as we know microelectronics can make them, all of a sudden the manufactures' market will disappear. For, apart from anything else, such emulators would essentially mean that any user buying the basic box (say at $500) plus Apple, Tandy and IBM emulator cards (at $1000 each) would effectively have the universal system from $3500.

The only option open to the manufactures would be, horror of horrors, built-in obsolescence and failure.

There is, of course, a good side to all this. Any degree of universality in hardware, even if it is only achieved indirectly through emulation modules, will allow software writers the opportunity of forgetting about the problem of which machine to write an application for. It will no longer matter too much - and it will also mean that once a program has been written and debugged, it won't have to be rewritten (and re-debugged) to run on another make of system. Universality of hardware will lead to the universality of software, and, as software is the real key to everything in this business, that can be no bad thing.

end