Main Content

Interviews


Click a photo to read one of the exclusive interviews

Martin Brennan


Martin Brennan

Martin was part of the team of engineers who formerly worked for Sinclair research and left to set up their own company called Flare technology. Flare technology whilst working for Sinclair research came up with the idea for a computer system code named Loki. It was this idea that they further developed into the Flare one after leaving Sinclair. When Wyn Holloway heard of the power of the Flare one and discovered that it was being offered for integration into existing products, he instantly jumped on board and together Flare and Konix evolved the 4 chip Flare one system into the Slipstream ASIC.
Martin and one of his Flare partners John Matheison off the back of the Slipstream ASIC took advantage of advances in technology and the lessons learnt from the Slipstream and developed the Flare two which Atari bought and used as the basis of their Jaguar console.

Martin went on to work with Wyn, Robert Kent and Michael Baxter again at MSU where his talents were put to good use delivering successfully products that initially found new uses for the Slipstream ASIC from the Multi-system and he also went on to develop a lot of the electronics that have found their way into home entertainment products that we all take for granted.

Interview:


Slipstream
What I'm trying to do is get to the bottom of why the Multi-system never got to market.

Martin Brennan
We designed the Flare one system out of our experience at Sinclair. The Flare one system was a demonstrator with 4 custom chips, we got that assembled, we got ourselves onto the cover of a magazine, I think ACE - we did some great demos - using not photographic quality graphics, but in those days anything was better than the PC's of the time, for example CGA with 16 colours. We had 256 colours so you could approximate photographs and we could animate them and have them sliding across and play music at the same time, we may have had a basic interpreter as well. And Wyn got in contact with us and offered to take it to the next step.
So he commissioned us to design a custom chip - maybe it was AMI or Toshiba who would make it... He also wanted us to use a different processor. We had used a Z80, but that's a 16 bit address range, he suggested we used the 8088 20 bit address range of the 8086.

Slipstream
So why would he specify the 8086 processor and not a more obvious gaming machine processor like the 68000 - It's not the most common processor for a games machine - in fact I only know of one range of machine that use that processor which is a Japanese range called the FM Towns.
I appreciate that it was almost irrelevant which processor you used anyway in the case of this architecture, as the customs chips were doing most of the real work.

Martin Brennan
I think that was our view, as we put it, delivering games performance is all about saturating memory - you've got a certain amount of bandwidth and if you can use it then you can't do any more and you couldn't saturate memory bandwidth for graphics using a traditional processor as they would fetch, then there would be several pipeline stages then they would fetch again so they were quiet for most of the time in actual fact. So we designed DSP's and blitters to cycle share and saturate bandwidth when they needed it so the CPU was the leader of the orchestra essentially so it didn't need to be that powerful.
It needed some more addressing range than the Z80 could offer but not a lot more, but the 8088 was great because it's an 8086 with an 8 bit bus externally so it's got all the same instructions. But I think it probably came down to price - this was to be a mass market product. It's not so much what it would cost to make - it's what the market would bear and where the manufacturers were pitching it, the 8088 was probably pitched at an 8bit market and therefore didn't have the premium of a 68000. We just lived and breathed low low cost in our Sinclair days, so we would never dream of using an unnecessarily exotic processor - it was always cut-throat so that would have been the reason for that choice.

Slipstream
There's a lot of talk on the internet about the Loki project - Can you tell me what that was and if it had anything to do with the Flare one project?

Martin Brennan
Loki was a concept - John Mathesion came up with the name. Loki was the Norse God of war I think. I worked on the Spectrum 128, It was a more reliable Spectrum - it's very boring, but I eliminated the -12 volt supply, using more up to date memory, we didn't need the -12 volt supply, therefore we eliminated an inverter, therefore we eliminated the biggest cause of failure of all Spectrums. In theory it was going to be a greater capacity with a greater reliability, we also enhanced the basic and added a sound chip. And there's the 128.
But we could tell where computers were going and what people would want. We wanted a real successor to the Spectrum as opposed the QL which was more 'businessy'. So we wanted something pretty much with the specification of Flare one but it was a concept and not a project - It had a name but there weren't any drawings as such. So when the company [Sinclair] was bought by Amstrad we thought that this is still a great idea, and we want to pursue it, and that we did.

Slipstream
So it was your intellectual property as Flare rather than Sinclair research's?

Martin Brennan
Obviously we couldn't use anything we had at Sinclair so we basically designed the "Flare one" for want of a better name from the ground up.
What you saw on the cover of ACE was the 4 custom chips, we had a prototype version that occupied 9 PCB's of TTL. It was a massive hand laid, wire-wrapped collection of printed circuit boards, that did a very similar job. It was completely un-portable - it occupied the whole size of a desk but we established the basic design on that and it took many months to build.
It was Charles Cotton who was a marketing manager at Sinclair or even a director, who said "people aren't going to believe that. It's too big and clumsy - you're going to have to make chips." That was one of the best pieces of advice we ever had in business, we just did enough consultancy work to pay the NRE charges to QDOS, a local Cambridge company - they did the chips for cheap because their software wasn't finished so we were sort of beta test users. So they got improved software and we got cheaper chips. They were essentially one-offs as they were direct writes - so there was no mask charge otherwise we could never have built those chips.
So it's derived from Loki, but as far as I remember Loki didn't ever really get anywhere.

Slipstream
From reading features in ACE, it seemed that the Flare one was ostensibly aimed at providing a very strong musical ability with it's built in MIDI capabilities and such - is that true?

Martin Brennan
We based a lot of it on the Amiga as well. That was a great influence - they seemed to have designed a very powerful games machine - if not a games computer then at least a games capable system. They had Commodores purse to design that - and it showed - there seemed to be several different ways you could do the same thing on the Amiga which struck us as overkill. We felt there were lots of instances of doing the same job - but one way - the best way, and it would end up being cheaper. Plus there were certain things we didn't like about the architecture - it was bit planes architecture rather than a bit per pixel, so we felt we had some ideas that would make our lives and the game writers lives that much easier.
When you say was the Flare one derived from the Loki - I think you've got to say that Loki and Flare one were derived from the Amiga, so you've got to pay tribute to those guys.

I was interested in sound sampling and felt that if you had a machine capable of complex graphics, then you could produce good sampled sounds and not just simple beeps etc.
I don't think the term "multimedia" had been coined at that time, but the Flare one design was capable of digitised images and sound and we saw that it had to have a DSP and a Blitter - two processors that completely outperform the central processor.

Ben Cheese was our music expert - he sadly died a few years back. He was a Musician and enjoyed the Saxophone and he pretty much steered us in that we must have a DSP and it must be a particular precision and be one of his recommended available architectures. He came up with the design for the DSP. It was pipelined to Harvard architecture and like the Blitter it saturated the memory bandwidth when it needed to.

Slipstream
There's a picture of you and Margaret Thatcher in one of the ACE articles, you were at QDOS. How did that occur?

Martin meets PM Margaret Thatcher
Martin meets PM Margaret Thatcher

Martin Brennan
My part in her downfall! She had come to visit Cambridge and among other places she came to visit QDOS and because we were one of their first and biggest customers, we were using four of the chips, and because we were producing chips for games machines - which was something other than for washing machines, we had the most photogenic chips, so I ended up on the photograph.

Slipstream
You looked like you had explained something really technical that had left her bewildered!

Martin Brennan
Well, everyone was incredibly nervous, there were security guards there - she was the Prime minister at the time, and there were protesters outside, so it was all pretty exciting.
She was great in that she put everybody at their ease - I think we had a laugh because I'd made some reference about what children would be doing in their bedrooms and she took it the wrong way.

Slipstream
Other than Konix had you been approached by any other manufacturers with an interest in the Flare one design at all?

Martin Brennan
Bellfruit took an interest and actually put it into many thousand of quiz machines.

Slipstream
Did they use it pretty much as per the Flare one design, or did they modify it at all to suit their needs?

Martin Brennan
They bought the chips from Texas instruments - once the design was proven, TI made masks that could then be mass produced. So they basically paid us a royalty. I can't remember who did the software for the machines though.

Slipstream
So where did the Atari Panther fit into the great scheme of things?

A Panther Kallista - Ultimately the reason the Atari Jaguar is so called
A Panther Kallista

Martin Brennan
Throughout that period we did consultancy work - we did work for Amstrad (we designed a fax machine for them and a hard disk controller). At one point one of the guys from Sinclair joined Atari. He had worked for Perihelion - Richard Miller. He became a director of Atari in Sunnyvale and he had a project called Panther - It wasn't called Panther when I joined. Panther was the name of the car my wife had just bought, a Panther Kallista and the chip had no name and I wanted to give it a handle - so it was called Panther.

The design and specification had already been started, and they said "somebody's left - here's the concept" and it was only the video part of the chip - there was no sound.
It was a novel video architecture that allowed you to create windows of different sizes and different bit depths. Essentially you didn't have a frame store - it was a composite of frame stores - a kind of smart video frame store. It would have allowed a great deal of sprite style animation. Sprites in general in those days would have been of a fixed size e.g. 16x16. The games looked 'spritey' because of that, this would have been quite an interesting departure. I wasn't keen on it, but I designed it and the chip was built.
But while I was over in California in '89, I actually convinced the bosses at Atari that 3D was the way to go, with the experience we'd gained on Flare one - if you didn't just do flat rendering, but shaded rendering you got a 3D appearance.
At the time, I was seeing pictures in magazines where computers were rendering photo realistic 3D wire meshes and I said "these are static images, but they only contain a very few number of polygons - we could take that, animate it and you could produce a game that was a quantum leap away from the current games".

So the Jaguar project was born from the Panther project.
In essence Atari looked at the Panther and looked at what we were promising for the Atari project and said can the Panther project.

The original design for the Jaguar was that it was actually going to be a 128bit computer, it wasn't going to be 64bit. We felt we had the pins to do it. We were going to have 2 banks of memory at 64bits and do double data rate and achieve a 128 bit architecture on 64 pins. It was really pushing it - but in the end the economics said that 128 wasn't necessary and it would have been too expensive.

Slipstream
Was Sinclair research your first job after graduation?

Martin Brennan
No it was my 4th - At the time the Spectrum can out, I wrote a game for it - Othello - or Reversi and made contact with them at that point. And at that point I was already doing consultancy which I found far more fun. I was working down in the south of France. I had difficulty getting payment out of somebody and I said to Sinclair "You know that job you offered me - how about it?" and they gave me a job.

Slipstream
Was that a commercial game that got sold?

Martin Brennan
Yes - it paid for a car.

Slipstream
That was back in the days you could do that - poor games programmers these days work ungodly hours for a relative pittance.

Martin Brennan
My wife works for a games company and you're right - I think it's like electronics; Engineers like what they do so they don't mind so much. I read something the other day that said that what you get paid depends on your skills, the rareness of the skills, but more importantly the unpleasantness of the work. If the job is unpleasant you will have to get paid more. It's the obverse for engineering - It's essentially almost therapeutic. It's actually very satisfying and good fun; people will do it for nothing.

Slipstream
It was an unusual pairing to put the Flare one electronics into Wyn's Slipstream controller - back in those days (and still today unfortunately) games machines are often aesthetically very boring.

Martin Brennan
Wyn realised that a better joystick would give you more feel - more immersion in the game, and I think he also recognised that you need a better computer for that. So to stick the computer in the joystick is where it all came about.
I think there's a good argument for it because if you don't package a sophisticated joystick with the computer, then the software writers don't have any reference other than the base - you know one or two buttons and a digital joystick. If the computer has a proportional controller and so many degrees of freedom then every game can be written to it so the games will be better, so it made a lot of sense to me.

Slipstream
That's certainly the approach used in writing games for a specific console these days as opposed to say a PC.

Martin Brennan
Certainly, you buy a PSP, you know the controller you're going to get.

Slipstream
That's the problem with writing games for PC, you have no idea what hardware the user may have so you can't take advantage of specific hardware. If for example only 3% of people have that hardware and 97% don't. Which is where abstraction layers such as Direct X prove their worth.

Martin Brennan
Of course you're right, I'd never thought of that.
I thought it was a great combination. I thought it had vision. We were always kind of arms length from the project - we'd done the chip design, and we had a vision of where the technology would go - but we equally had his concept which was driven from the market place and was about making money. I think at that time in our lives, yes, we wanted to make money to pay bills, but we weren't so driven by money - we got a lot of satisfaction from doing it - probably too much satisfaction.

Slipstream
Once you'd fulfilled your contract with Konix to provide the technology - how involved were you in revisions and modifications to the design such as the increase in memory capacity suggested by the games programmers?

Martin Brennan
I can't remember how many revisions of the Slipstream chip there were. I don't remember any disasters - When you design a chip it's like writing a program, it's very complicated - but you don't get to try it, you get to make it at the cost of £100,000 pounds - and you then ask yourself about the bugs that are there, are they serious, are they so trivial that can you work around them, are they serious and there would be a loss of functionality that could stop you selling it. Or are they so bad that you'd have to respin the chip. I'm not sure we hand any type 3 situations, or even type 2 - we had a lot of niggles. I know that it has been respun a few times, and the last one would take it into MSU which was no longer Konix - but much the same people. And they used it in a web browser with a 386.

Slipstream
Was the deal with Konix an exclusive one? Could you have sold it to others once Konix took it on?

Martin Brennan
The chip that they had was exclusive to them - I seem to remember a discussion with them that went along the lines of "We want exclusivity - how much will it cost?" You can have exclusive - but it'll cost you about 3 times what you're paying and they decided they didn't want it at that price. It was effectively exclusive anyway because we didn't sell the chip to anybody else.

Slipstream
Would you have made revenue from every chip sold?

Martin Brennan
Yes.

Slipstream
So that must have been quite galling for you to have missed out on that potential income?

Martin Brennan
Absolutely - I remember there were stage payments - and the final stage payments took a lot of dragging out of them. I'm sure there were stage payments still owed to Ben Cheese even after he had formed his own consultancy company and when John had gone to Atari. Well after the Jaguar had hit the streets - I was still getting payments for the design work they had done years before.

Slipstream
Wyn has led me to believe that there were external forces that deemed the Multi-system such a threat that financial backing that was in place (or promised) was being withdrawn and thus making it even more difficult to fund the project. What's your take on this?

Martin Brennan
Wyn really has to be credited with having a great skill for knocking down doors - for being able to establish communications with factories and retailers and getting things done.

I could say more about Jaguar.

Around the time of the Summer spring CES I remember it was at the same time as the World Cup. Sony announced that they were going to spend 200 million dollars acquiring software for the Playstation. They'd hired Chicago's rock café and invited software developers to come to it.

When you have pockets that deep you're not going to fail. I happen to know that at that time the Atari family had made 60 million dollars cash from the sale of some land in Taiwan that they had judiciously purchased.
60 million dollars is a great deal of money for a family - but they must have had to consider putting their 60 million up against Sony's 200 million dollars - I'm not sure I'd have made any other choice than theirs. You can't compete with that - even if you have a better product or not.

Slipstream
Building up to the intended launch of the Multi-system, it looked like the KMS was going to dominate the entire global market.

Martin Brennan
I think we thought it was a foregone conclusion at that stage - the chip did the job, the software was coming along and the plastic we assumed was there.

Slipstream
Do you think the Flare one would have better placed as a home micro rather than a games machine?

Martin Brennan
You have to ask why did the Spectrum do so well? People bought the Spectrum because it had a look. When you get down to it, it was a damn good looking thing. We just assumed as Flare that if you put great hardware in a box then that would sell - but as I've gotten older I've come to realise it's how it looks.

Slipstream
I suppose you only need to look at the Ipod. The MP3 player I'm recording this conversation on is technically far superior to an Ipod - but I'm a geek, so I'm more interested in technical side of things - but the Ipod is so successful and it really doesn't matter what's inside as long as it does the job. People who aren't interested in electronics just want a device that's easy to use, does the job and looks good.

Martin Brennan
Great example - you just have to look at the Playstation - the PS2 I mean, it was a good look - but now that vertical orientation is starting to look dated.

Slipstream
They've refined the design for the PS3 - but it's still capable of being orientated vertically just like the PS2 was.

Martin Brennan
Yes, it's like the monolith from 2001: A Space Odyssey.

Slipstream
What sort of performance targets did you have for the Flare one / Multi-system?

Martin Brennan
We thought that there were certain components around and that if put together in the right way, there was an increased performance you got from putting them together in an elegant way. There was nothing kludged, we basically said we'll take this memory chip, we're going to design a custom set of processors to go with it and the TV video standards defined the quality of the video.
We just thought that certain things fell out very naturally unlike the Amiga. With it's bit planes architecture.
And CD quality audio was the benchmark to try to hit.

Slipstream
There were some concerns about the design of the Flare one - where you could either have great graphics, or great sound, but not both at the same time because of the contention.

Martin Brennan
I'm not aware of those concerns - one of the weaknesses of our business plan was that we designed the chip and a demo and left it at that, we left the developers to get on with it. It was a tough job for the games programmers.
I think we should have produced an SDK to make programmers lives easier, this may have addressed any concerns that people had.

Slipstream
Did you ever meet Jeff Minter? He had produced a great looking game AMC '89 which looked like it was pushing the machines technology well.

Martin Brennan
Yes - once. He seemed to be able to produce work far more rapidly than just about anybody else.

Slipstream
He has left his mark on many cutting edge machines, from the KMS to the Jaguar and the Xbox360 via the Nuon which Richard Miller and John Matheison worked on. He was involved in the design of that and had written a game and some other stuff for it.

Martin Brennan
Oh did he? Yes he was doing work for Jaguar. Is he still writing games? I must look him up.

Slipstream
Yes he's still very active - his latest project I believe is Space Giraffe soon to be released on the Xbox360 and he has a column in Edge.

Slipstream
The Jaguar had nicknames for its chipset - Tom and Jerry. Did the Slipstream have any chipset nicknames?

Martin Brennan
I think we just called it SS1.

Slipstream
Have you any idea why Atari liked to use cat names as the basis for its product names?

Martin Brennan
Well - I came up with the name Panther as it was my wife's car. Sam Tramiel gave it the name Jaguar - he thought it was an English car name, so he gave it the name Jaguar.

I remember he gave me a test drive in his brand new Ferrari - they were all car fanatics in that family.

Slipstream
You were involved with Wyn Holloway again in MSU - where you developed a prototype games machine based on Slipstream for a Chinese and a Taiwanese company.

Martin Brennan
Yes, we took the Slipstream IP and developed it to a 386 architecture. And they did a couple of things with that.
They wanted to do a web browser and I think they were pretty ahead of the game. They had a TV based web browser. This was just when the internet had really only just started up.

John and I were given the choice of what to do - John went to work for Atari and I worked full time for MSU.

I also designed a CD controller chip there. We realised CD chipsets cost 7 dollars so I thought we could easily do the same job for only 3.50 dollars but that could be sold for 7 dollars. But the market was 100 million! And it was a move away from these speculative 'gamesey' type things where there's no existing market into a more stable market.

MSU was floated on NASDAQ and I made a great deal of money because as a key member of the company at the start I was given the option to buy shares, and I exercised that option and sold them on NASDAQ and made quite a lot of money.

Slipstream
How long did you spend at MSU?

Martin Brennan
It must have been 4 years.

After MSU I formed the Cheap and Cheerful Chip Company.

Slipstream
Was that its real name?

Martin Brennan
Yes.

Slipstream
Excellent!

Martin Brennan
There's a heritage in that - there's a company called QDOS - quick and dirty operating system - I thought that was a good handle.
At that time every body was getting into 3D graphics and it was getting saturated, I liked the idea of going into audio. I'd had a broken hi-fi at home, so I opened it up to fix the switch and I saw that it had 14 chips in it, so using the same logic I thought I could do this with just one chip.
So I raised 2 million pounds in venture capital and the company changed its name to Global Silicon. I sold out a few years back but the company is still going.

4 years ago, I set up my current company based on the simple idea that a hard disk is a great way to store your music collection - and not much more than that.

Martin Brennan
There was a guy at Konix who subsequently moved on to MSU, I know for a fact that he had exactly the same shares as I did, he actually left. I was thinking about this, this morning, I read a book about NASDAQ and about stocks and the quality of stocks that had been traded back in the 90's, I realized that applied to the shares I made money on.
But this guy sold his earlier than I did. He sold them earlier than I did by virtue of the fact that he left the company, you weren't allowed to sell them if you were still in the company as you had a vesting period or something. But because he left, he was then free to sell them.

So he sold them at a higher price than I did. My shares were worth 6 million dollars. And I grimly watched as the value of my shares went down and down. I think he may have got close to that figure when he sold. I can't remember what his surname was. He was a PR agent if you can trace him down then he'd have some interesting stories to tell.

Slipstream
Oh, god what was his surname, I've got it here somewhere [I grovel around in my folder to find the copy of Zero magazine with the article that mentions the name].
I remember because there was an article in Zero that made it seem like a large quantity of the Konix machine had been manufactured and was waiting in storage - It was probably just PR nonsense.

Martin Brennan
That's the ting we were as much consumers of those stories as everyone else unfortunately.

Have you spoken to Robert Kent?

Slipstream
I haven't got a good lead on him yet, but I'm actually getting a bit overloaded by all the information and contacts to talk to at the moment.

[I find the article in Zero] Michael Baxter of Solution PR!

Martin Brennan
Now track him down because he'll have a different perspective because he was quite close. So much of this is about creating Hype.

Slipstream
[I pull out the Slipstream reference manual and show it to Martin]
It's amazing what you can buy on eBay these days isn't it!

Martin Brennan
Good God! I've probably got the source of this somewhere, amazing - and its got the DSP instruction set.

Slipstream
If only there was some source code then someone could use this to aid in producing an emulator.

Martin Brennan
Sheesh, It's a lot of work isn't it…

Slipstream
Yes, but there are some dedicated people out there.

Martin Brennan
So this, based on the date from the copyright, this is just prior to me going out to California the year after in 1989. So that would have been completed and I went of to start working with Atari.
So, why didn't Wyn make it with that - to be honest it's the end of the business that I'm only just turning to 20 years later. You know if he could have sold it to people, the next step was to produce in small quantities - maybe hundreds for people to try - I didn't see them so I think the money was too short to get to that stage.
He wouldn't have sold it out to anyone that's for sure.

Slipstream
There was a very definite deal that almost went through with Lucasarts. With talk of Sony manufacturing the equipment - which is quite significant as it was prior to Sony building the SNES CD add-on and would have been Sony's first involvement in making games machines - which has obvious implications for Playstation etc.

Martin Brennan
Now I do feel sad - that would have bought more than a car, a mansion and a yacht for me!

Slipstream
In hindsight you'd have though Wyn should have sold it.

Martin Brennan
It's hard to overestimate Wyn's drive and ego - I don't mean that in a bad way, he got so much done by it.

Slipstream
What are your views on the games industry now as a designer of gaming hardware back in those early days? There seems to be a lot of power being thrown at the games in terms of huge processors.

Martin Brennan
From an engineering perspective - that stuff is entirely predictable. It was quite obvious to me that there was far more processing power than was needed for business so the best outlet for that was going to be entertainment.

When we look back, I don't know whether boxes like the PS2 were ultimately the best place for that power - but we'll look back and see them as a niche, at point solution.

Where it's going right now, a game takes a team of over 100, so Flare couldn't have come into being, so it's no longer 3 guys in a garage that created Flare one.

I think you have to look at something like Shrek where the artists have said we're going to model an individual single speck of dust or hair going across the female leads face.

This will all come into the games, and the realism will just improve and improve and improve, and because it's all being delivered by transistors, yes you can put it all in a box, but I think it's all going to come back to where is it created - where's the creativity coming from. I think that'll become the bottle neck.

It's more a question of where do we look for ideas and not just how much memory do we need or how much processing power.

I said years ago, On demand is the only game in town. And it can be generated globally - meaning that technologically once you can provide video on demand by virtue of the internet than a lot of the other questions become rather moot. All you'll need is a TV - it's just a window on the world. Then you can move away from the individual boxes.

You can then focus on book writers and artists for ground breaking content. The imagination will become the premium.

I think it'll move completely away from technology into artistry.

Martin is still producing cutting edge electronic devices and products at his company 3GA. You can find his Biography on his company website at http://www.brennan.co.uk