Interview - Chris Curry of Acorn - PCN

ACORN COMPUTER, best known as maker of the BBC Micro, is fast becoming a major force in computing. Next year turnover will probably exceed £30 million and the company plans a major assault on the American market.

Compared to its Cambridge-based rival Sinclair Research, Acorn does not gener- ate the same almost fawning coverage in the computing and general press, and the personalities behind the company are cer- tainly less well known. Chris Curry, who with Hermann Hauser founded Acorn Computer, does not find this at all sur- prising.

"Apple is much more our competitor than Sinclair. We have a fairly wide base compared with Sinclairís monolithic, pro- duct-based approach, so we will never make such dramatic impacts. Clive brings out one product and pushes that straight . towards whatever sector of the market is appropriate. Generally speaking for him it is the consumer market. Whereas ours is a company with diverse interests, hav- ing customers in the consumer, develop- ment, education and office sectors."

Chris Curry worked for Clive Sinclair for 13 years, first in hi-fi at Sinclair Radionics, then calculators, and finally computers at Science of Cambridge, which Chris Curry set up for Sinclair. "To start with I was the only person there. We

Chris Curry

Acorn is a company to be reconed with in
computing, yet te forces behind it are not so well
known. Ian Sobie meets Chris Curry, one of Acorn's
founding fathers.

were producing a little gadget, a wrist calculator, a ghastly thing. I- used to do everything: placing the ads; packing kits; taking them to the post office; mending some of them; answering technical queries; absolutely everything. Later on there was a secretary, then more and more people were involved. By the time the MK14 came out I think there were about five or six people there."

The MK14 was one of the very first cheap computers. It cost £39.95 in kit form and came with 256 bytes of RAM, a hex keypad, an eight-digit display, and was based around the National Semicon- ductors four-bit SC/MP chip.

"It was quite a successful computer. The next step was obviously a version that' ran Basic instead of just machine code. That was where our ways sepa-

rated, because Clive didnít want to do it and I did. So I set up Acorn Computer with Hermann Hauser."

Their first product was the Acorn Sys- tem One, which came out in January 1978. "We chose the word Acorn because it was going to be an expanding and growth-oriented system." The System One was a semi-professional 6502 de- velopment system aimed at engineering and laboratory users, but priced low .enough, at around £80, to appeal also to the more serious enthusiast.

One thing that stands out about Acorn is how consistently the company has stuck with the philosophy that the design must generally take into account future de- velopments. By the same token a new Acorn product usually has some conti- nuity with its predecessors.

"We still use parts of what you could call the System One, the Eurocard sys- tem, as our rack-mounted file server for Econet systems. The Basic that went into the Atom was a slight variation of the Basic that was written for the System One, a fast, control-oriented Basic." The System One established the 6502 as Acornís standard processor, to be used later in the Atom and BBC machine, and the Torch and Electron.

Hermann Hauser and Chris Curry met through Cambridge University while Curry was still at Science of Cambridge, and the university connection has been very valuable to Acorn.

Glittering prizes

"We are in Cambridge because the university processor group, or its compu- ter lab, has maintained Cambridge as one of the leading universities in computer science. So there is a good supply of bright people to go into the industry. One of our directors is a lecturer at the com- puter lab, and he is on the watch for good people about to leave that might want to work for us. And of course an awful lot of hobbyists come out with software that they have written in their spare time to sell us, and we make contact that way. Cambridge is a fairly small town; people know each other. There are lots of small software houses, all closely interlinked with the university, and we all talk to each other. To move out of this area would be a major disadvantage."

Acornís current products are: the Atom; the BBC Microcomputer model A

62 PRACTICAL COMPUTING October 1982


of Acorn

America in the Autumn when the Elec-tron comes out. The English-speaking countries, Australia, New Zealand, South Africa, are our best bets as all the course material and software is im-mediately appropriate. But I have no intention of just limiting it to those mar-kets. We have already made arrange-ments with people to translate all our documentation into Italian and Spanish."

The Acorn philosophy is to look well ahead, but not lightly to throw away tried and tested systems. So it comes as no surprise that Chris Curry intends the Atom to have a continued life even when the Electron becomes available.

"The beauty of the Atom is that you can get inside it, you can tack bits on to it, you can use it as part of another, larger, piece of kit. The Electron will be a fairly finite system, and because it will have, like the BBC machines, ULAs and so on, it will be very closed to an electronics engineer who might want to use it. The Atom is still a fine machine for tampering with hardware."

Long-term strategy

The Gluon is an interesting system as it uses the 32-bit processor chip from National Semiconductors, the 16032. Superficially this looks like an aggressive attempt by Acorn to leapfrog the com- petition, who are mostly migrating from eight-bit processors to 16-bit chips, prin- cipally the Motorola 68000. But Chris Curryís explanation shows that the usual long-term and even slightly defensive Acorn strategy is being followed in this case as well.

"We chose to go with National Semi- conductors, who we have been working with very closely, rather than with the 68000, Z-8000 or 8086, some time ago. I think we have made quite a good choice, because there are already many people on the bandwagon who we could be competing with if we went towards the 68000 or one of the others. But we are in . with a head start with the 16032, which happens to outperform the others pretty heavily anyway. National Semiconduc- tors have set up a package de'al called the EP scheme, which ties in a lot of suppliers of software and hardware to provide the chip with the support it needs. At the end of this year we will be able to offer a whole range of languages and programming tools, all that you need to start developing systems."

The Gluon comes in two versions, one intended as an add-on box for the BBC machine, the second as an add-on for the popular commonly available, eight-bit microcomputers. The BBC Gluon comes with 0.25Mbyte of RAM and the 16032 processor. It sits in a box beside the BBC machine connected by the Tube, the very high speed data highway used in the BBC machine which gives the system much of its long-term expandability. In effect the (continued on page 69)

and B; and Econet, Acornís low-cost networking system. A division of Acorn called Orbis deals with the high- performance networking system, the Cambridge Ring. In addition Acorn sup- plies Torch Computers with boards to go in the office-oriented Torch machine. Acorn helped set up Torch, and Chris Curry was a director, but this link has now been severed and Torch is supplied on a straightforward OEM basis.

A new version of the rack-mounted development system, called the System Five, is due soon, continuing the System One tradition. But the major new pro- ducts from Acorn are the consumer- oriented low-cost Electron, and the top- end add-on processor system, the Gluon.

"The Electron is designed to compete with the Spectrum. The idea is to get the starting price very low, but not preclude expansion in the long term. It runs the same Basic as the BBC machine. We anticipate it having just an extension bus, and the extension bus will have modules plugged into it to give you whatever interfaces you want. So an Econet inter-

face would be a plug-in module, a printer interface would be a plug-in module." The Electron is due out in the late autumn. The price is not yet announced but it will probably be under 1200 to compete with the Vic and Dragon, as well as the Sinclair Spectrum.

Chris Curry clearly thinks that BBC Basic has a good chance of becoming a world standard. "An awful lot of work has gone into that language, almost to the extent of having committees deciding things. The Electron will be very useful for people to get in at a lower cost level, and gradually more and more people will adopt our language."

Acorn plans to start a major assault on overseas markets with the Electron and BBC machines. "We are actually doing very little exporting at the moment be- cause of the backlog on the domestic market. Our production levels are very high indeed, and when we have cleared it we will be producing far more than we are selling in the U.K., so we will be making a very heavy push overseas. There will be very heavy advertising in

PRACTICAL COMPUTING October 1982 63

(continued from page 63)
6502 in the BBC machine becomes a front-end I/0 processor and all the lan- guage progressing is done in the 16032.

Version two, the Universal Gluon, cannot use the Tube because it is unique to the RBC machine, but will connect via some serial or parallel interface, probably

initially the RS-232. It will be a bigger package, comprising a 16032, up to 1Mbyte of RAM, and 1-, 2- or 5Mbyte Winchester hard-disc drives. It comes with terminal emulation software, so in effect your Apple, Tandy or Pet becomes just a terminal connected to the Gluon 32-bit computer, where the real action is.

The Gluon will be making its appear- ance, vaguely scheduled far the end of this year, at the same time as major computer companies such as DEC and IBM enter the U.K. microcomputer mar- ket. Part of the appeal is the very high standard of the display and keyboard, and this may well encourage the raising of ergonomic standards generally.

The Gluon, using say a Tandy or an Apple as a terminal, will have imposed on it the ergonic limitations of the older machine. But Chris Curry does not think this will limit the Gluonís appeal.

"Apple and Tandy owners want to maintain the usefulness of their equip- ment, they donít necessarily want to change over completely to a 32-bit computer. Remember that a very large proportion of the existing installed user base of things like Apples are in the hands of home users, and those people donít want to lash out enormously on a completely new system

This version of Gluon is best under- stood as a migration tool, intended to pick people up from other manufacturersí existing user base and move totem on to Acorn kit. The price has not been announced yet, but it is clearly crucial to the success of the strategy.

Acorn will be offering two different operating systems far the Gluon; its own very minimal operating system, and a version of Unix, "We always see ourselves as supplying development- hardware for people in the systems house business, arid that is what these small operating systems are for. It also gives us an intimacy with the product that a lot of other people who buy in system software do not have. To do serious commercial applications software people can go for

the more expensive option - Unix.

Many of the new 16-bit machines from other manufacturers run Unix, which has a good chance of Incoming the standard 16-bit operating system in the same way CP/M has become the eight-bit standard. To the computer user, and more so to the programmer; the operating system de- fines a computer system, usually much . more strongly than any hardware feature. So the Gluon will be another Unix- system as far as the potential buyer is concerned, and probably one of the cheapest.

What price performance?

Chris Curry believes Acorn has another advantage. "There have already been some 68000 add-ons provided for the Apple so we know we are competing with people using the other 16-bit proces- sors. But because we know that our one, with the 16032, outperforms them, we feel that we have got a fairly clear tech- nical edge."

Practical Computing has been a bastion of seepticism as to how far the perform- ance of the chip at the heart of a system actually concerns the end-user, once all

the other hardware and software factors' have been taken into account. Systems are rarely CPU-bound. So how far does Chris Curry think the fact that the 16032 outperforms the 68000 really matter?

"Probably not as much as one can imply. It is a matter of presentation, advertising. You can say this will run three times as fast as a 68000. Now whether or not that is desperately import- ant in real terms I donít know, I am not a user myself; I know that if I went up to the labs and asked someone that theyíd say of course the speed of operation is vitally important, It means you can do things which you wouldnít be able to do otherwise. This certainly is the case when it is part of a total system, with its own screen, doing its own graphics. Where it is on the end of ay RS-232 perhaps it is a little less important. Nevertheless, I think it appeals to people, that they are getting the latest, in that terrible phrase, ĎState of the Artí."

Both versions of the Gluon, and the Electron, which is really a cut down BBC machine priced to appeal to the consumer market, demonstrates the flexibility Acornís systems approach gives them.

"Because of the way we have designed things it is particularly easy for us to bring out a machine for a particular market. We can both spread them out and add more facilities, or we can prune them down dramatically to produce a much cheaper machine."

It is one reason why Chris Curry does not fear a price war, which seems more and mare likely as new companies enter the increasingly defined market and pro- duction capacity builds up. "If we need to go to a lower price we bring out a new product that costs us less, we donít cut prices."

One of the most successful recent com- puters is the Osborne 1, a cheap portable computer with everything, including screen and discs, in a single carrying-case. Given its flexibility, why had Acorn not done it first?

"We thought about it, and always con- sidered that the sort of small screen that you could put in the thing was not really practical. What we did not really expect was that the very notion of a complete box with everything in it would be so attractive. And it obviously has been because it seems to be very successful. Mind you we are aware of its success and there are products in the pipeline."

Take-away Electron

The likely date for Acorn to bring out its first portable machine is June 1983. "It won't look much like an Osborne but it will be a machine that includes its own display facilities. Because it will have limited interfaces and it must be as small as possible, it will be based on the Elec- tron rather than the BBC machine. It will have a very strong emphasis on com- munications. If it is used in an office it will expect to see a local file station acting as its storage. It will have an Econet local area network interface and Modems for the British Telecom network;"

Success brings with it new problems. Appleís success has made it the target for piracy, in the form of cheep imitation machines which pass themselves off as

Apples, and legitimate competition from manufacturers of plug-compatible machines such as the Dutch Pearcom and German Basis 108. They are intended to be sufficiently similar to the Apple to accept all hardware and software add-ons yet use different circuit boards and have (continued on page 71)

PRACTICAL COMPUTING October 1982 69
(continued from page 69)

some significant additional features.

The BBC machine would seem a temp- ting target for copiers if it proves as successful as Chris Curry hopes. "You would have a job to do it with the BBC machine because it uses ULAs, and these are not so easy to copy If you have a

design the main blueprint of which is the printed-circuit board, then this gives the circuit away and there is no way of pro- tecting it. The point about the Apple is that it is all standard components on the PCB so it is a prime opportunity for somebody to copy."

The circuit is burned into ULAs late in the production process and cannot be discovered by physical examination. The functions of the circuit would need to be analysed and a similar circuit designed to carry them out - a lengthy process. "That is why Sinclair can be relatively safe with the ZX-81. It is the perfect thing to copy in Taiwan, but for the fact it has ULAs in it."

Acorn has been expanding fast. "We donít make the sort of profits that Sinclair makes but we are putting a lot more back into the company." With expansion has come reorganisation. "We started out with a lot of bright young chaps up at university doing design work, and found that we were, as we always intended to be, heavily research and design oriented. But with substantial production we have found that there is quite a lot of manage- ment to be done. So quite recently we have installed a financial controller, a software manager, and a chap in charge of Orbis Computers to look after net- working.

"One of the things that we have to expand enormously is our marketing abil- ity. The Sinclair type of marketing approach is almost entirely mail order, plus one mass merchandiser. It is very simple to organise with a minimum of staff. We are now going to go into having mail order, mass merchandisers, deal- erships, and direct institutional sales on quite a large scale. So our marketing side is being expanded very dramatically."

By the meagre standards of a high- technology computer company Acorn is becoming a major employer. Most of this is in design, marketing and administra- tion. "Out of our total workforce of about 75 people only about eight or nine are actively putting things together. We

do not do any assembly work ourselves except the final configuration of systems. There are four production lines making products for us in the U.K.: ICL at Kidsgrove; Race Electronics and AB Electronics in South Wales; and RH Electronics in Cambridge - about another 220 people."

Dangerous business

Acorn also has a Hong Kong supplier. "They make 50 percent of the Atoms that we build at the moment, and they are also a source of components from the Far East. We have known them a long time and can trust them not to go into Taiwan for any of those awful copies they make. We have to provide detailed listings of the operating system software to our sub- contractors to enable them to do testing properly, and that is a fairly dangerous business, especially if it is going over- seas."

The main reason for assembly in Hong Kong is to supply the Far East market. "The cost of shipping becomes a very

significant factor. Completed units are fairly bulky, which is why we are not desparately opposed in the long term to local assembly overseas where it is desir- able on political grounds. If we can get just casing and final packaging done over- seas, and ship tested boards in bulk packs, that would mean the majority of the work, in value terms, is done in the U.K."

Acorn does come up against barriers abroad. "There is a strong requirement for local assembly in an awful lot of countries; certainly in America. We have not started marketing in North America yet, but when we do we will find that in a very short time it will be in our interests to do at least some local assembly and deal very closely with an American com- pany. In South America it is very largely required that local, or at least part local, assembly is done. Europe is OK. France is of course extremely partisan; it is diffi- cult to sell anything into France, at least in the education field."

On the other hand Acorn has been the beneficiary of U.K. government policy, in particular the Department of Industry scheme to put one micro in every primary

school. The machine had to be British. Chris Curry does not think this has helped Acorn much: "I am a free market person and absolutely against all forms of protectionism. The Americans practise it against U.K. products, incidentally, so I suppose it is only fair that there is a little bit done here. But is it beneficial? The education market for us has been quite a small proportion of our business; so far it hasnít made a lot of difference to Acorn."

The market Acorn is in would be a difficult one for an entrepreneur in any country. It is changing rapidly, with very large companies moving into it with new products and attempting to force the mar- ket towards paths that suit them. As the market stabilises around a predictable range of defined products the massive electronics production facilities in the Far East can be brought to bear on particular, attractive market segments.

Not like calculators

"If the consumer end of the personal- computer market ends up like the calcu- lator market then I donít suppose we would bother to be in it, because there wonít be any reward. I am not sure whether it will, because the software aspect of personal computers does pro- tect it from the Far East onslaught to quite a large extent."

Apple has been making predictions of a major crisis in 1984, when all the curves on its graphs lead to an expectation of massive overproduction. Only five or six companies are expected to survive with any appreciable market presence.

Acorn emerged in the early days of small computers, the days of a general excess of demand over supply. It was a period typified by tiny companies, long waiting lists for products and forgiving, enthusiastic consumers. Chris Curryís

strategy for Acorn in the coming brave new world of computing is to concentrate on building an organisation sufficiently large and sufficiently flexible to survive.

"For the next year we will grow as much as our profits will allow us to grow - wide growth in all directions that we are technically capable of going into. Then we will select whatever sphere of operation is the best bet for us. I donít think just one, but we will select perhaps three areas to concentrate on. We will survive by having lots of strings to our bow." µ

PRACTICAL COMPUTING October 1982 69