• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Stop wasting time looking for files and revisions. Connect your Gmail, DriveDropbox, and Slack accounts and in less than 2 minutes, Dokkio will automatically organize all your file attachments. Learn more and claim your free account.

View
 

Computer Equipment for Display and Data Processing

Page history last edited by Alan Hartley-Smith 7 months, 1 week ago

D&DH

Home

Introduction

As users of radar systems demanded more and improved performance the development of digital computers provided the means for automatic plot extraction and processing of data to enable better display and control for operators. Marconi designed both special-purpose and general-purpose machines using first discrete transistors and then microelectronic circuits to meet this need.

 

With the formation of Computer and Automation Divisions the use spread to other industrial applications as well. The details of these divisions and their capabilities are shown here.

 

 

 

From an Elliott-referenced source:

 

1959 - BTM + Powers-Samas = ICT
1961          + GEC computer interests = ICT
1962          + EMI computer interests = ICT
1963          + Ferranti EDP computer interests = ICT


1963 - English Electric computer interests + Leo computers = EEL
1964          + Marconi computer interests = EELM
                 + Elliott Automation = EEC

1968 - ICT + EEC = ICL

 

A related History- LEO

 

 

Just a reminder of the early days - how things have changed!

 

 

 

 

input from Steve Bousfield

Elliott’s real time machines stayed in the Marconi fold i.e didn’t go into EELM and evolved through a Marconi Elliott Computers;  GEC Computers route. At least two parts of GEC-Marconi had considerable investment in software for the GEC (formerly Elliott) 900 series range of computers. When the last of these, the 905, was being run down with no further manufacture planned by GEC Computers they took matters into their own hands and designed their own replacements.

Firstly Marconi Avionics designed the 920ATC  (Advanced Technology Computer). It was, to all intents and purposes an updated 905 which they needed to support their airborne computers which had been sold in large numbers to the U.S.
Then Marconi Defence Systems designed the MC1800 which was a microprocessor based emulator of the 905.
 
Radar, which also had a lot of 900 series software, used both of these as 905 replacements in its Radar Simulators with 920ATCs going to, among others the Royal Australian Air Force and the MC1800 to the RAF

 

Input by Roy Simons

IMP and TAC and Locus 16 and Myriad were all started and implemented whilst they were in the Display and Data Handling labs and when the same people moved to Computer Division they continued with further product development.

 

Editors note - lecture to the North West branch of the Computer Conservation Society

The PowerPoint file is available here. The content is quite full but it proved acceptable to the computer-literate audience and evolved into more of a dialogue followed by a detailed question period, and I believe has established the Marconi contribution to computing within the Society at least. My thanks are due to all those involved in the original work and who helped in putting the presentation together.

 

TAC

Don Ward - I know of three TAC installation, Wylfa, FUR HAT and Malvern. Three computers or more?

Peter Bain - Nassau Phase I had TACs - I don't know how many - I expect Cyril would.

Steve Kay - Wylfa had two and I understand FUR HAT had three.

 

Running total - 7?

 

 

Panic

Don Ward - Three PANICs at each PD site, nine in all?

 

Running total - 9

 

Mentor

This was derivative machine but the design was overtaken by events and became dated so was never completed. There are some papers in the collection held in the Manchester University archive.

 

Myriad

(including its progenitor IMP ('Ispeed Miniature Processor)

 

ATC at Bretigny

 

Don Ward - SAUDI had 6 Myriads, how many on NASSAU was it one Peter, I really should know that! How many on FPPS? Any other sites?

Bernard de Neumann - three at Baddow

Steve Bousfield - the Cambridge telescope array used Myriad to keep its dishes aligned, didn’t London Transport have one as well? The Australians also had Myriad as part of their HUBCAP Air Defence System (They had two – they were transportable and the D&DH was Decca / Plessey using Myriads but whose radars they were I don’t know. Addendum by Roy Simons - On 1 Apr 68, 114 Squadron reformed at Amberley equipped with the "Hubcap" Air Defence System. Re-equipped with the AN/TPS-43 radar in 1979, 114 then operated as a highly mobile Control and Reporting Post, developing the deployment techniques and battlefield operating procedures that have established 114's reputation as a tactical unit. Addendum by Steve - the TPS43 radars used Westinghouse D&DH systems. They were part of Australian Project RECAP which was the successor to HUBCAP. We supplied the simulators for Project RECAP.

Don Halstead - I think one of the first production Myriads went to CERL (Central Electricity Research Labs) at Leatherhead.

Ian Gillis - The Project Nassau SCC had a redundant Myriad computer system with two 32K (24-bit word) Myriads for intercept control and weapon assessment, an 8K for inter-trace marking and a 4K for auto-follow, each SRS had a Myriad. Two MOV's were provided, each containing a Myriad computer. I think that makes 9 in total. Addendum by Peter Bain - The Myriads at each of the three SRSs were 4K if I remember correctly, the Display computer I thought was 16K rather than 8K. The MOVs had 16K Myriads I believe. 

Peter Bain - Nassau Phase II had 7 Myriads. Nassau MOVs had 2. I believe FPPS had 3. There were also Myriads on Tonic & Venture. There was a Myriad III at Malvern for RRE (as it was then). Plessey bought a number of Myriads for fitting in vehicles for the Australian armed forces. I believe Mayflower had a Myriad for Sweden? Wasn't there a Myriad controlling a steel making process in Wales?

ANO - At least three Myriad (probably 2s) at West Drayton for FPPS. I believe one Myriad 3 was sold to somewhere in RRE in the late 70s. Line Division used Myriad as part of MARS (Marconi Automatic Relay System) - sold to Nicosia in Cyprus for AFTN work. NATO. Defence Communications Centre London. Space Division used Myriad for Satellite Tracking. Southampton University Sound and Vibration Lab. Deutsche Forschungsanstatt fur Luft-und-Raumfahrt. Central Electricity Research Laboratory. Wylfa Head power station ? UK Hospital. Road Traffic Control - Glasgow, West London. LATCC. Computer Aided Design using Myriad and X2000 displays; X2000 displays- Midland Electricity Board, British Steel Corp, Pembroke and Drax and Hinckley Point power stations.

Mike Steeds - Myriad I; Add GEC research labs Stafford; CERL [Central Electricity Reseach] Leatherhead; RRL Glasgow [ Road research Labs.]; France[ I think Thompson CSF] - Myriad II; NHS Royal Hospital Belfast [Heart rhythm monitoring]; Cambridge University [Star Gazing]; Steel Co. of Wales [ Production of steel, Port Talbot]; Capetown University - Myriad III; Used for programming at WRW Bureau. London Transport had a Myriad II to control the No.11 bus route installed on the platform at Mansion house. GLC also had a MyriadII to control London traffic lights in West Central London.

Don Ward - As I was responsible for the software for Saudi, Venezuela, Pakistan and Iran I should know. I am not sure of Venezuela, maybe 2 Myriad 1s, Colin Birch would know he went there. Pakistan and Iran had Mobiles with 16k Myriad Is but how many I don't know. There must be records somewhere.

Colin Birch - In your tally of computers note that Saudi was a Myriad 1 system but several Locus 16s were installed some years later, interfacing to the Myriads to provide multi-radar capabilities at the SOCs on Project Simcats which was an ATC system with a bit of military on the side.

Colin Birch - In all the emails I have seen no mention of the four Myriad IIs sold to Comms for the Defence Communications Centre project in Whitehall.  They were used in a main/standby configuration, one pair for communications handling and the other pair for database.  These were interfaced to CDC disc drives and development work was done by Radar to achieve this.  Len Hope spent some time in Whitehall getting it all working and I think Rex Manville was involved with the disc interface.  This was all in the early 70s.

Peter Hopp - The Message Switching Group at Writtle under Florian Lipinski, (latterly the Computer System Division) used Myriads as Comms Processors from about 1966 when I joined them in their first incarnation for their Military and Civil Aviation Message Switching Systems, I will have a further think about how many, but my initial memory is:
1966/7? NATO Everrre Brussels 2 x Myriad 1
1968? MET Office Bracknell 2 x Myriad 2 (there may even have been a second system later)
1969? MOD Northwood 2 X Myriad 2
1970? MOD DCC Whitehall 4 x Myriad 2
1970 ATC Cyprus 2 x Myriad 2
197? Chile Santiago MSS
To support the development team we had at least 2 x Myriad 1 and then probably 2 x Myriad 2 in our Bureau at Writtle before they were replaced (amid much toil and tribulation - the company wanted us to have GEC 4000 machinery!) by a Vax 11/780 and then a second I believe. We then went onto using our own design of multi-microprocessor based systems for a whole range of applications across the whole company product lines. I will think further, particularly dates, and also check with others

 

Running total - 58

 

FPPS

See Barry Jones's note on the Myriad system and its evolution on the FPPS page.

 

Locus 16

 

 

Approximate numbers for the following systems are indicated below. Further Locus quantities for workshops would have been delivered:

 

20 - Furnace systems (based on 2 Locus bins per furnace cabin (or demounted equivalent) and 10 cabins); TBA - MACE systems; 20 - TOR; 50 - ScATCC incl radar processors, display processors, workshop etc (probably a few more); 10 - Queen Alia ATCC; TBA - 713; 20 - KAWAL; 30 - Astrids (used a special programmed Locus bin, 1 for each display); 30 - MRSL Bureau, labs, workshops; 2 - Warton; 2 - CAA Radar Evaluation System (PEST) vehicle.

Other Locus systems were delivered for Tepigen (one variant used Locus, another used DEC computer), naval (small patrol boat) systems and possibly other systems

 

50 - UKADGE (this might be closer to a per site qty if so perhaps a factor of 5-10 out. Addendum by Peter Bain - There were over 200 UKADGE consoles with a Locus per console. Addendum by George Duncan - On UKADGE you had 41 Locus consoles at Buchan, Boulmer, Ash and Neatishead then 4 consoles at Portreath, Saxton Wold, Bishops Court, Benbecula, Bentley Priory and High Wycombe not to forget the Faroes. During life of system consoles were also at Cosford, West Drayton and the software facility at BP. Remember the LDET and RDET. The only two to exist now are a console at Neatishead and Manston both devoid of inners. Have been to all sites. Our test and repair equipment moved Writtle Road works, Chelmsford; Blackbird Road works and Scudamore Road works, Leicester; Broadoak works, Portsmouth two consoles. Locus still lives.);

 

ScATCC

 

Input from George Duncan - More on Locus 16 there were at least forty 19 inch racks at Prestwick Scottish air traffic control centre. Locus was used on S600 S605 S713,723,743 800 there were over 70 types of PCB. Most have now been removed and replaced.

 

Input from Derek Knight - there was also a mobile system called PEST - Plot Extractor System Test- that was a lorry with a Locus and 3017-10 display with re-timed video. Supplied circa 1978 from memory to the CAA that they drove around their radar sites to check on the alignment between video and plot data. This was when the use of synthetic only displays (ie ScatCC) were becoming operational and operators needed convincing that all was well.

 

Ed - I came across this item in my trawls - an interview with Seb Welford

 

Running total - current inputs indicate approaching 600 and counting - if correct this quantity must rank high in industry deliveries.

 

Useful computer reference sites.

 

Input by Bernard de Neumann

I recently acquired a book: "Alan Turing and His Contemporaries - Building the world's first computers", edited by Simon Lavington, published by the British Computer Society, 2012.

Whilst it covers the period up to the late 1960s but with only brief notes for the last 10 years of that period, there is very little mention of Marconi computers, especially 'IMP, TAC, MYRIAD, and LOCUS, despite the fact that two TACs ran in tandem to control Wylfa Nuclear Power Station continuously from 1968 - 2004, having been installed in 1966! Thus they ran for 36 years non-stop! Is this a record? Anyway, it seems to me that the effort to record Marconi Radar history should include something on these computers, and that it should receive wider publicity to ensure that historians of computer history do not overlook them in the future.

 

Addendum

I've been in touch with Simon Lavington, author of the book I mentioned in my post, and although he said he had heard of TAC and MYRIAD, he had never heard of 'IMP.  Simon is the Computer Conservation Society's digital Archivist and Professor Emeritus of Computing at the University of Essex, so whatever we come up with, in my opinion, a copy should go to him for their archive. I've also found that he is author of "Moving Targets: Elliot Automation and the Dawn of the Computer Age in Britain, 1947 - 67", Springer, 2011. It's quite a big book, running to over 700 pages.  Whilst I haven't access to the book, on Amazon there is a "look inside" facility that mentions the following Marconi computers: M2100 series, MYRIAD I, MYRIAD II, and TAC.  There is also mention of "System 4" (some of the smaller ones of which were manufactured by Marconi, but that is not mentioned in the index). [Editors note - Brian Partridge says "The System 4/30 was designed for EELM by Marconi Computer Division at Chelmsford using techniques developed for the Myriad computer. Unlike a Myriad it was a computer for commercial not real-time use". See diagram above. There is still a persistent mistaken view of Chelmsford involvement in the EELM work - see here]

 

Comment by David Samways - Brian Partridge is correct - the System 4/30 was designed by Marconi for EELM to market.  The other System 4 models, which were based on the RCA Spectra 70 series, were developed by EELM at Kidsgrove.

 

 

Input from Alan Mathews

I was not directly involved with computer developments at Marconi but was mostly working with those who were.

For IMP/Myriad - Don Becket did the core store and the late Derek Jeffries most of the logic, with Eric Atkins aiding and abetting this slightly unofficial development in the early stages. I think Andy King and Phil Goulden may also have been involved somewhere. IMP was alleged at one time to be the fastest computer in the world !!!

I helped Bob Donaldson doing tests on TAC in 1966/67 to set up the stores for temperature stability using "smoo" diagrams in Room 100 at Baddow and applying this to the Myriads we sold for the South African MOV contract.

I seem to remember that Rex Manville designed a disc store system for the Myriad for Wylfa Head ?


Others will no doubt know much more about these developments than I do and Brian Partridge must have been heavily involved before his later deep involvement with Locus.

 

Input from Brian Partridge

Just to confirm. With IMP the ferrite store and arithmetic logic were done as stated but the microprograms implementing the instructions were done by myself with assistance on the algorithms from Digby Worthy. The Project was funded to do research on storage techniques and fast logic using early microcircuits. Senior management were not informed that we were designing a full blooded computer until we showed them it working. Soon after that we were commissioned to make Myriad using the same team. One significant aim was that it should be a computer in a desk rather that an array of racks like all computers at that time.

It was also considered important that it should survive in operation for short power breaks which meant that power was supplied by batteries which were kept charged during power breaks. During the development of the battery system we had the frightening experience of batteries exploding in a run away situation which we could not stop and we had to sit on the floor behind benches to avoid injury for over one hour.!!!!

 

Input from Don Ward

Very interesting discussion on Marconi Computers. I didn't stay long enough to really use the LOCUS 16. I think it's first use would be for Yugoslavia with Peter Bain et al.


One machine seldom mentioned was PANIC, used on the PD system. A TAC was installed at Malvern for the trial PD system . PANIC was (I think) a development from TAC but was reduced from 20 bit to 18 bit words. The magnetic core store was also reduced from TACs massive(!) 4096 words to 2048 words. This was the RRE's idea to reduce the cost! It was designed specifically for PD and never used anywhere else. The design team included Derek Canfield and Brian Mellor and others. The mind boggles at what we did with a 16K Myriad compared with today's memory.

 

Input from Alan Mathews

We were told at a recent lecture on car electronics, that the chip in each door handle on a modern car used 32K of storage. I pointed out that we used to get the software for a complete air defence system into this amount of store, including tracking, aircraft performance data and display management etc.etc. !!

The lecturer also stated that some cars now had more lines of code than a Typhoon Fighter. So things now must either be bloody sophisticated or the code very inefficient - I suspect the latter.

 

Input from Alan Mathews

Roy Simons is as usual quite correct, but Brian Partridge probably will have the best technical information on these Marconi computer developments.

As I think I wrote earlier, the IMP was developed in Eric Atkin's lab at Baddow in a slightly clandestine way as a research project into high speed ferrite stores by Don Beckett and high speed logic by Derek Jeffries. They joined the two developments together and hey-presto there was a computer - possibly the fastest in the world. It was nicknamed IMP and is now stored, I understand, by the Science Museum and formed the design basis for the MYRIAD.

There were several Myriad versions - the first ones with a very elegant Teak case which looked like an expensive executive's desk and these were sold to South Africa and other customers. Later more powerful versions were in metal cabinets.

The Honeywell H112 however was the central "bought-in" machine used for the first S600 computer based mobile radar data handling cabin named FORGE - intended to be a next generation after the ANVIL cabin. It had alpha numeric display characters and rate aided tracking facilities etc. This design I remember also formed the basis of an ATC system which was sold to Malaysia, but which used a YAMATAKI computer to replace the H112 which had gone out of production. Norman Davies will know all about this system, because he was the RME there !!

Forge was offered to Yugoslavia but they wanted a more powerful system, so the H112 was replaced by Locus 16 for this contract to provide auto tracking, airfield status and other features demanded by the customer.

I do not think we ever sold a FORGE S600 mobile cabin but the development of the hardware and software led to other products which were successful.

Hope I have got most of the "facts" right.

 

Input from Norman Davies

I don't believe the Honeywell H112 based operations cabin was ever sold, but a demounted version of it formed the basis of the ATC display systems for Malaysia. Two systems were installed over the period 1974 - 1976, the first at Subang (Kuala Lumpur ACC) and the other at Senai airport (Jahore Baru). Each system consisted of two sets of equipment, interlinked to allow interconsole pointing between the displays on either processor. The Subang system used two Honeywell H112 computers, but for the Senai system we had to use H112s made by by Honeywell-Yamataki. These had cosmetic differences, but were otherwise identical.

 

I believe the H112 was primarily used as a process controller and was used in the oil industry, so it presumably had a pretty good environmetal spec.. It was certainly reliable, I am struggling to remember a single failure of the Honeywell H112s at Subang over its' operational life.  (It even survived the fire which destroyed the Subang ACC and was able to be redeployed in the temporary facility we set up.)

 

The H112 had a single bit arithmetic and logic unit so carried out serial operations on the 12 bit words. The H112 on the Malaysian systems had 8K of store (the maximum) of which 4K was used for the application, 2K for the display buffer and 2K spare. (The 2K spare capacity was intended to cater for an SSR upgrade.) The direct addressing range was limited to 128 words. I am sure this proved something of a challenge for programming. (Ask Colin Birch and Chris Devine.)

 

Input from Colin Birch

The H112 was used in a FORGE cabin funded on PV.  It was programmed to interface to a Myriad to provide the operational functionality of the Saudi system but the H112 provided a touch interface to the operator.  A working demonstrator was produced but it was never sold to anyone.  With a little more research I could perhaps come up with dates, but it was probably around 1971.  Much of the operational requirement work was done by Bob Prior, an ex-RAF controller, who worked on the Saudi system, and beyond FORGE, carried the work through to FURNACE.

 

Although FORGE as a cabin was never sold, the equipment configuration (1 H112 driving three displays whose IDs Matty or Norman Davies will know) was sold to Malaysia for use at Subang International Airport along with an ATC radar for en-route control.  The system provided six controller positions using two H112s talking to each other and apparently it ran continuously for 25 years until a fire in the terminal building damaged it.  This was installed in the early 70s and extended about 5 years later to interface to a new Raytheon Approach Control system.  A second system was installed under Project Genting with a radar in the Genting highlands but this used Yamatake DC12s which were carbon copies of the now out-of-production H112.  Norman knows much more detail as he looked after it all on site for many years.

 

I also worked on FURNACE systems sold to Yugoslavia (Projects Bridge and Bridget) and Oman but perhaps I can say more about those another time.

 

I noted that Don Ward suggested that I went to Venezuela on Project Venture, but in fact it was Saudi where I spent some time commissioning the software there on the five SOCs and the ADC.  Venture was never installed and commissioned – the radar rusted away and some of the data handling equipment was dropped off dock when being unloaded.  I can’t remember how many sites were involved but I think there was an ADOC and at probably one SOC. 

 

In all the MOGS messages I have not seen any mention of Project IBAC, an ATC system for Iran.  Ask Geoff Wheeler about it – he was in charge.  There was also an ATC system for Oman, quite separate from the military stuff.

 

Input by Steve Bousfield

Two of the projects that came over from Elliotts were the Plot Extractors for IBAC (3 off) and Norway. Alan Cheesewright was the Systems Engineer on IBAC (Irano British Airports Consortium) if I remember rightly.

Alan Batchelor ran the Forge System with H112s in a ‘lab’ in E Block (at the end of the first corridor on the left as you came in the door opposite D Block.)

 

Jordan had a string of stuff, some things that I remember, though without the detail are:
S511 and displays for Queen Alia (included a Solartron ATC Simulator)
S711 – the only ‘real’ S711s sold as far as I’m aware (i.e with the hydraulic mast as opposed to the Turkish ones that were hand cranked)
S600 convoy (donated by Saudi)
Long Range Air Defence Radars (can’t remember the type) at least one of which was retrofitted with the ‘big valve’ transmitter designed for the S713 Martello, thereby giving rise to the stories that Martello had been supplied to Jordan which was not true.

 

Various contributions to Discussions on Computer Notes on MOGS

The Myriad was developed as a consequence of engineering prototyping of ideas. These were the IMP (hI speed Minature Processor) that began as an experiment into high speed logic design and ferrite core storage. It led to Myriad 1 that evolved to Myriad 2 and then in the very late 60s/70s post FPPS to Myriad 3. EE (then the parent of Marconi) was sold to GEC in 1968. It took some time for the reorganisation to be achieved and create the product companies (GECC, MRSL etc). GEC policy determined that the only part of the company making computers was GEC Computers. Myriad 3 was the planned computing backbone for Forge cabins. GECC’s computers were not rugged enough for MRSL’s needs [Editors note - opinions differ on this point - see previous input from Norman Davis]. A replacement to Myriad was sought. H112 was chosen [Editors note - any specific reason for this choice] which led to or extended the Forge interfaces.

 

The Honeywell H112 was to be used as a display processor interfacing with the PPI & tabular displays. The central processor to which the H112 interfaced was the Myriad 3.

 

It is believed a Forge demo was created by Don Ward using Myriad software from the Saudi system. A PV project for software development for Forge was started but did not get beyond defining the tote formats required. Forge got overtaken by Furnace & the Bridge & Bridget contracts. The software PV work was then used as the definition of the Bridge totes.

 

Honeywell ceased production of the H112 and MRSLs investment in hardware interfaces etc. and software etc was all wasted - we might have reused interfaces in later designs such as the Furnace cabins which were Locus 16 based yet had the older 3017-03 or 3015 displays fitted; not the later 3017-10 Locus 16 type and there were some odd oldish interfaces. The important part of this aspect is the relatively uncontrollable costs of depending on another supplier whose rational decisions in their world could threaten projects perhaps even the whole company. This could have led to the decision to go in-house to Locus 16 rather than just not-invented-here/politics [Editors note - was this the case?]. [Subsequent amendment - when the H112 became unavailable it was replaced by a Locus 16 with a special interface board giving the same interface to the PPIs & tabs. The Myriad 3 was also replaced by a second Locus within the ops cabin. Forge which consisted of a Computing Cabin plus Ops Cabins thus evolved into Furnace with Ops Cabins which also contained the computing.]

 

Increased miniaturisation allowed the concept of local processing associated with each display console in place of the centralised concepts and the problems with system reliability exposed by the FPPS system. MRSL began such a development based around a data route (subsequently known as a bus or highway). Space only allowed for 16 bits and so compatibility with Myriad wasn’t possible.

 

There were advantages in the Locus 16 instruction set as it allowed fully re-entrant code, ability to have multiple central processors designed-in at the start etc. which Myriad couldn’t do. But that meant starting to build software support and libraries etc. from scratch again, plus extensive training, all at some cost.

 

The early policy was to not develop on Locus 16 itself (linked to GECC policy as data-processor development by business units were tolerated whereas computers were software building machines and they had to be GECC) which meant that original support was on the Myriad bureau at Baddow, later on the Myriad 3 and GEC bureau at WRW and then later still on the Vax bureau. Such decisions impacted the cost of providing bureaux etc. to customers and the very clunky, even then, paper tape based and FLEXDOS compilation system on Locus 16 itself.

 

Also the computer policy that still determined that only GEC Computers could be called computers existed but a radar data processor was OK being just further down the chain than a signal processor and OK for unit design. 

 

The original product name was intended to be Data Route but another product already used this and a new name of Locus 16 was established. The sale to CAA established this product.

 

 

D&DH

Home

Comments (3)

Ian Gillis said

at 6:24 pm on Feb 10, 2016

Page checked - dead RAAF linked removed

barryjones@... said

at 1:16 pm on Feb 8, 2018

The FPPS operational system at West Drayton comprised 3 Myriad 1s each with 32K of 24 bit memory and a 4K external store (based on the X2000 system graphics memory) as a sort of high-speed cache. A fourth similar unit was used as a development bureau.

When, in 1970, it was discovered that this configuration was not large or powerful enough to do the complete civil and military task, a series of studies were undertaken to work out a solution. There were some advocates for the use of Myriad 3 (still a glint in the designers' eyes) but as there didn't seem to have been much thought about the software required for the 15-bit addressing system to handle more than 32K of memory, the working group was unable to recommend its use.

The system was descoped to only handle the military task and the Ministry, whose air traffic control responsibilities were shortly to be devolved to the CAA, initiated the studies that resulted in the acquisition of the American IBM 360-based 9020 system as the core Flight Data Processing System with consolation awards to UK industry of a data entry system to Marconi and a radar data processing system to Plessey.

The former contract was terminated after system definition in the belief that passing notes to a central keyboarding service would be adequate. Experience showed that this optimism was unfounded and Myriad peripherals were reintroduced into the civil operations room as the "Interim Sector Update" system. This part of the system stayed in service until 1991 although the military Flight Plan Processing System had been retired two years earlier.

Ian Gillis said

at 3:04 pm on Feb 8, 2018

Thanks for your contribution Barry - I've added it to the FPPS page at http://marconiradarhistory.pbworks.com/w/page/32770586/FPPS

You don't have permission to comment on this page.