Chris Shelton is not well known today, yet the British microcomputer industry would have been a very much poorer place without him. Never as famous as Sir Clive Sinclair, with whom he worked in the past; Acorn’s Chris Curry, Herman Hauser, Steve Furber and Sophie Wilson; or even Tangerine and Oric’s Paul Johnson. Nonetheless, …
You can sort of see the problem with the UK industry, always trying to go against the grain and not use the Intel processor.
If Sinclair and others had just used Intel and focussed on producing some cutting edge software then they might still be in the computer industry.
Too many people focus on the nuts and bolts, the hardware, when all that really does it run the software.
"I live in the future, not the past"
I hope to be saying that when I am in my mid 70s.
Only one possible icon...
One of the interesting points made was that the UK enthusiasts had a lot less spending power than their left pondian counterparts, to put it simply : this means that they could not afford the same gear as the left pondians.
A Z80 clone was much cheaper than an x86 in the 80s, I doubt it was any different in the mid-late 70s, and the uP+RAM would make up a very big chunk of the budget back in the 70s.
When what they were trying to do
Was run the software on cheaper, faster processors then no, the basic idea wasn't flawed. As usual for British companies it was their attempts to market it that failed. If they HAD managed to sell the design to a manufacturer then the resulting machines would have beaten Intel to a pulp, being both cheaper to produce and significantly faster. What it looks like they were lacking when it came to their sales pitch was a software stack (compilers, OS etc) that would alow simple porting of the IBM/MSDOS world's main packages and some sort of agreement in principle from the software houses to produce versions for it.
Was there enough info available to design with the 8086 at the time the NASCOM was designed? Not to mention the '86 was released when the NASCOM had been on sale for some time. I assumed AC 09:18 was talking about the 8080, which I'd have thought would be an unlikely choice with Z80 being an improved 8080 as I understand it. Your cost point about RAM and CPU is obviously a prime consideration for the intention of the computer, but I suspect board design would have been more complex and costly too - I recall reading around the time the IBM PC came out that the design team had gone for Intel out of familiarity with the 8080/its support chips and chosen the 8088 specifically to simplify the board. Definitely agree about the spending power thing - still seems to be much the same with the pricing of US stuff here.
Back in the 70s and early 80s Intel was not the powerhouse you know today. Zilog and Motorola processors were more popular and common than Intel. Like MS, Intel owes a lot of it's current position to IBM.
I think the only flaw in any of this was the marketing. Offering the chip to the entrenched incumbents. Don't suppose they be overjoyed by anything so "disruptive". Can't help thinking that if Sir C had slipped the chip into something himself - just something updated-QL-esque even - the rest of the industry would have soon found itself clamouring to buy/licence the chips and catch up. How different it all could have been.
Also wondering, esp. reading p4, if this wasn't the inspiration behind Intel's radical "innovation" of the Pentium Pro with its fast RISC core and CISC interpreter. Timing would have been spot on.
Sad to see that the first comment on here is from some AC who understands nothing about the UK home computer boom of the 80s.
The UK never had the (military) budget that the US had to pour into the nascent micro-electronics sector.
But what it lacked in raw cash it certainly made up for in hardware and software innovation.
After burning down a few 8080's (being on nMOS, it absolutely needed -5V to be applied before +5 and +12) it was high time to ditch it and replace with Z80. Could even double the clock then.
So yes, Z80 was better than 8080. Alas not pin-compatible.
"Too many people focus on the nuts and bolts, the hardware, when all that really does it run the software."
Seems fair to me.
"If Sinclair and others focussed on producing some cutting edge software then they might still be in the computer industry."
"If Sinclair and others had just used Intel then they might still be in the computer industry."
It's Intel that are increasingly irrelevant, in the post-PC era.
You can still buy Motorola 68K for use in e.g. harsh environment embedded systems where re-engineering the existing 68k system would be a nightmare.
You can buy something called ARM (architected in the UK) for almost everything else apart from the IT departments. There, x86 still has a role in datacentres and desktops. But for how much longer?
Re: Too polite
"t the inspiration behind Intel's radical "innovation" of the Pentium Pro with its fast RISC core and CISC interpreter. "
Might be. Then again, the records show that in the mid/late 1990s, a company called DIGITAL sued Intel on the basis that the Pentium Pro had been infringing patents relating to a variety of subjects (branch prediction, cache coherence, etc). A couple of DIGITAL chip architects had arrived at Intel shortly prior to that (this was around the time that first implementations of the 64bit Alpha chips, capable of running UNIX, NT, or OpenVMS, had been arriving on the market).
As part of the eventual settlement, DIGITAL's CEO was allegedly offered a secret deal where Intel would offer DIGITAL preferential access to the IA64 technology, in exchange for which DIGITAL would have to take Alpha off the market, leaving the 64bit "industry standard" to be dominated by IA64.
Unfortunately for Intel, AMD had different ideas about who was going to lead the "industry standard 64bit" race.
Re: Too polite
In fairness I think the concept of a sort of front end that prefetched, grouped up and dispatched instructions was already history, IBM flogged it in the form of STRETCH in 1961. The neat thing about that PgC widget for me is the asynchronous core throttled by temperature. I am guessing in practice the upper limit would be the I/O pads. Would be fun to fab something like that today, I would want to throw in lots and lots of ECC protection. :)
Re: Too polite
Anonymous Coward Spaffed Thusly:
"Also wondering, esp. reading p4, if this wasn't the inspiration behind Intel's radical "innovation" of the Pentium Pro with its fast RISC core and CISC interpreter. Timing would have been spot on."
John Cocke developed the IBM 801, the very first RISC in 1975, and that design was evolved to drive various I/O processors right through to the 9370 in 1986 (which was an 801 derived core pretending to be a S/370).
Jeez, I think I'm going mad, I have plugged IBM architectures as class leaders twice in one day...
Re: Too polite
>John Cocke developed the IBM 801, the very first RISC in 1975, and that design was evolved to drive various I/O processors right through to the 9370 in 1986 (which was an 801 derived core pretending to be a S/370).
Weren't they synchronous though?
(It's *thus* not "thusly" btw)
Re: Too polite @ YAAC
"Weren't they synchronous though?"
Well, I would guess that the IBM cores were synchronous in the sense that they were probably clocked, but that is orthogonal to the similarity I was drawing between the two (ie: the instruction execution is decoupled from instruction fetch with a view to keeping the core busy while waiting for main memory).
As it happens I didn't get the impression from the article that the PcG core used 'self-clocked' logic, because I was expecting it to have run with a cycle time that was an integer ratio of Q-Cache cycle time (2.5ns the stated SRAM cycle time - which fits nicely with the stated target of 200MIPS).
Interestingly DEC's 21064 (aka EV4) sampled at 150MHz in Feb '92, so the PcG was pretty much on a par with the Alpha in terms of timing - although I think it's fair to point out that the Alpha was 64bit, had floating point, cache, but it did cost quite a bit more... If you believe Wikipedia; DEC were punting 150MHz parts at ~$450 a year later.
Seems very cheap when I look back on it when I think back to how a 150MHz 21064A showed a PPro 200 clean set of heels running an identical WinZIP binary using the FX!32 x86 emulation layer (under NT 3.51 in 1996).
I suspect that the proposed 16-bit variant of the Sig-Net would have been earmarked to use Software 2000's TurboDOS. This was a multi-user operating system which borrowed a lot from CP/M, and which ran on the Intel 8086 series of processors.
I briefly supported a legacy TurboDOS system back in 1990. The main unit featured on 8086 processor, a hard drive, an S100 bus, a dumb terminal, and various printer ports. Into the S100 could be plugged a numbe of 'slave' boards - these each supported two extra users, via two 8086 processors and two dumb terminal ports.
"...always trying to go against the grain and not use the Intel processor"
Which is what Acorn did, leading to ARM. That was a bad move. Not.
I thought I was pretty good at obscure processor families but the Pgc was a new one on me.
Maybe they would of been better off targeting niche application areas. Defence and DSP spring to mind. The quantities would be smaller but the mark-up higher and there would be less competition from the likes of Intel.
Ironically considering that the defence monolith GEC plessy were actually making the things you would of thought they had a ready market there for things like sonar and radar applications within that organisation. Then again GEC was always technologically conservative.
It was also nice to be reminded of Anamartic. I remember at the time thinking solid state disks were just round the corner, not realizing there would be a 20 year delay. Maybe a good idea for the next article...
The Nascom was popular as the basis of various things that we'd probably classify as embedded systems nowadays. My favourite was the Movement Drum Computer, a sampled and synthesis based programmable drum machine. If you Google it you'll find it was used by Eurythmics amongst others, and an early model appeared in one of their videos (Sweet Dreams I think) with Dave Stewart tapping away at its keyboard.
managed to persuade the nice guys & gals in the MRSL training department that we needed a Nascom-1 in 1978. We installed it in the wonderful 'haunted' Springfield Place Grade-2 listed hostel next to the G8FKI ham system & antique broadcast TV cameras. It was a decent machine, tho' we eventually burnt out a PIO port playing music - but before that it did space invaders and other relevant entertainments of the time.
the students who played on the Nascom - some are now professors, one student even started what is now Amazon.co.uk, some obviously can't be named for legal reasons - I like to think our Nascom was helpful in all this...
Never seen details of the PgC7000, Would have liked a diagram.
They also had a rather novel bipolar gate circuit IIRC. It seemed to use deliberate latchup as a memory process.
Note that would have made it the second Bipolar microprocessor that Ferranti fabricated.
Too bad I don't think anything that good came along until the AMULET clockless ARM version.
Re: Never seen details of the PgC7000, Would have liked a diagram.
So would I, but Chris no longer has any documentation, and I've not been able to track it down. If any Reg readers *do* possess PcG documentation, please pass it on to share. Email me through the byline link in the main article, please.
Re: Never seen details of the PgC7000, Would have liked a diagram.
I would love to see some docs on that beast too.
Happy memories - and perspective
Nice to see Nascom being recognised as the UK's first microcomputer (I'm not sure "arguably" comes into it - if there was an earlier commercially available option with full keyboard/screen, then I don't know of it). I was a little late into the game as I started with the Nascom II.
Many years ago I designed and built a control and timing system for a car racing track using the Gemini system, which was basically a company setup by some Nascom staff when the company went into receivership (eventually bought by Lucas). The Gemini basically adopted the NAS-BUS renamed as 80-BUS and, for a while, there were compatible cards produced which would work with either Gemini or Lucas-Nascom systems. Sadly, whilst the race track was built and the system installed, the business was fundamentally flawed and now operates as a kart track. However, I still have my original Nascom II and the prototyped Gemini-based timing system, complete with the multi-tasking real time control system I designed for it.
As for the know-it-all who condemns blames the failure of Nascom on not choosing Intel, this shows an enormous amount of ignorance. At the time the Nascom was produced, the 8086/8088 family wasn't available (the Z80 was an improved 8080), and many manufacturers in both the US and the UK used other chips. Indeed the industry-standard OS at the time was CP/M and that was very much centred on the Z80 & 8080 architectures. At the time the Nascom was designed it could easily be said that it was based around what looked like an industry standard processor.
Meanwhile, many other companies used the Z80, some Motorola 68000 and yet others the MOS technology 6502, including one called Apple. Shame they came to nothing because they adopted the wrong chip for their first computer...
Of course the 8086 family only become the industry standard through marketing and the happy fact (coincidental or not) that IBM chose it for the basis of their belated entry into the business PC market thereby inadvertently setting an industry standard (in IBM's rush to jump aboard a boat that they very nearly missed, they weren't able to do what they'd done in the past and tied up designs with IBM proprietary standards - by the time they realised it, and sought to regain control - remember Micro Channel - it was too late). If it wasn't for this action by IBM the standard PC architecture might well have been based on another processor family. Indeed, in many ways, the Intel 8086 was one of the least likely choices on purely technical grounds - it was a flawed architecture, and arguable several of the alternatives at the time were better. For that matter, we may have had a different OS model - MS-DOS was little more than a CP/M look-alike for the 8086 family, and it's well known that IBM's adoption of the OS could easily have gone another way. It is not always the case that the "best" solutions win through in industry. Much also relies on luck or seemingly arbitrary decisions by people who just happen to be in the right position at the right time.
Of course there were many UK computer companies that did seek to adopt the Intel 8086 family and produce industry standard designs, and they have pretty well all disappeared as they got squashed by competition. The problem with just producing industry standard designs is that there's little room for true innovation. It becomes a commodity market, and the spoils go to those with the lowest cost base.
Finally, one few UK survivors in the computer market is ARM, the offspring of Acorn Computers, and they did innovate. For the record Acorn Computers did not adopt Intel processors, they started with the 6502.
An inspiration for a generation
I attended the Wembley show and remember it being beyond my wildest dreams!
I bought my Nascom 1 from Henrys Radio on Edgeware Road for just shy of £200 - just before finishing my A Levels. I remember being driven back to Birmingham and pouring over the manuals and writing snatches of code on random bits of paper I could find. I had built it that evening and lashed together a power supply - it work first time!
Having now programmed for too many years, my favorite is still Z80 machine code!
I have a lot to thank Chris Shelton for.
Re: An inspiration for a generation
Nascom, Henry's Radio, Edgeware Road - Ahhh. The memories!
Re: An inspiration for a generation
Edgware Rd - yes, what a memory. Not just Henry's but all the other component shops that opened up within a couple of hundred yards south of Henry's. And Maplin when it first opned up a few stores outside Southend (one in the back streets of Bromley as I remember) when they pretty much just did components. Proops for rumadging around? Those were the days!
Very cool. My first exposure to the inside of computers and using them as more than a black box was at a local computer club, where a older member called Jerry had a kit built Nascom2. I was fascinated that one week he'd slide in a new card and do a graphic of a pixel bloke running, another something else. It was housed in this old tv case. It led me to build a keyboard from scratch for my then brand new zx81, add ram to shunt the char rom into ram so it could be edited, and all sorts of other nasty hacks which were incredibly educational.
Much later on I found he'd swapped it for a telephone answering machine for his business, I'd have loved to have owned it, it'd be set up next to my amiga and things in the second office :(
I still warm a soldering iron up and tinker with embedded stuff now and again for fun.
Ah The Memories
Just had to go and dig out my Issue 1 of Personal Computer World featuring the Nascom 1 on the cover.
Imagine my surprise to find a Nascom 1 programming Manual alongside. I must have sent off for this - it is very comprehensive .
The Nascom 1 was too expensive for me - this was early 1978 and eventually when Sinclair brought out his ZX80, I said to SWMBO when they are £100 I will buy one. The ZX81 hit that pricemark and many happy hours programming late into the night followed!
I was a regular subscription reader of PCW and it was fun until the focus shifted to business machines and the software was costing more than the hardware. Flicking through these early PCWs there were some evocative names : the Exidy Sorcerer, Altair, Research Machines 380Z. PCW August 78 featured an Apple II review. That machine was a stupendous £1250 for a 16k machine only. Complete cost including TV and cassette player was £1520 +VAT.
Explains why homebrew and kit machines were so popular and I am sure a lot of readers on here cut their teeth on such systems.
Fun looking back through the very early PCWs - brought back some memories.
I understand that Dr Shelton is in fact 67, not quite in his 70s yet!
We had a Nascom-1 at school, in a very elegant wooden box as I recall. One Nascom amongst about 5 BBC Micros at the time.
I seem to remember some very enterprising and more talented and older than I writing a sideways scroller game for it. Dave C? Do you remember anything about that?
It took time for people to become tired with Intel and its ways. So when ARM came along it was a breath of fresh air.
RISC OS/ARM programmers were much in demand back then.
But it takes software not hardware to win success. Lyon wasn't the only company who didn't understand that people buy hardware to run software that's available. Not the other way round. :)
- Vid Hubble 'scope snaps 200,000-ton chunky crumble conundrum
- Bugger the jetpack, where's my 21st-century Psion?
- Google offers up its own Googlers in cloud channel chumship trawl
- Windows 8.1 Update 1 spewed online a MONTH early – by Microsoft
- Interview Global Warming IS REAL, argues sceptic mathematician - it just isn't THERMAGEDDON