|Home > LEO Computers > LEOPEDIA > Other Memoirs, Reminiscences > Anthony Robin Davies: Commentary|
Anthony Robin Davies: Commentary
I appreciate how your work must start with LEO I, but as one of the later development engineers (I was #24) I should like to see the whole story recorded, right through to the end. Effectively this end was on 16 December 1968, when all development work at Minerva Road was stopped and all of us had to either transfer to other ICL locations or be made redundant.
At the time the I/O control logic of the System 4/60 (nicknamed LEO 4 at the lab, and with architecture by me) was working and Maurice Blackburn's ALU was not far behind. Finishing it would have given ICL a machine that could really compete with IBM's System 3/60, rather than their feeble copy (the ICL 1900) of an antique Hewlett Johnson design. By then Maurice and I were wrestling with the interconnection problems of the "fail-soft multiprocessor" which was to become the System 4/65, 4/45 and 4/35, all of which contained aspects of LEO design thinking. And then there was the amazing sequel, when the extra III/F machines, chunks of which had been earlier redesigned by Tony Williams and myself for speed reasons, were produced for the GPO. In between there was a string of patents and innovations, including microprogramming using ROM, plug-in variable
storage speeds, 8-way multiprogramming, designing to the GPO work unit, novel and complex machine instructions like "merge, condense and sort", and CLEO (much of which resurfaced in the design of IBM's PL/1).
I imagine few people, when they plug a memory upgrade into the motherboard of their PC, know that the original concept, documented by Maurice and myself as co-inventors, was patented by ICL. Plus the breathtaking systems availability figure of over 99% for LEO III (contemporaneous machines from IBM only achieved 95%: IBM never disclosed this kind of figure but somewhere in my 25 years of employment with them I learned it).
On both the technical and social side it was a fabulous place to work: very friendly, supportive, empathetically managed, and without any visible sign of a rat-race. We were breaking new ground time and time again with the result that at the end I found that I had become the youngest Senior Principal Engineer in the history of ICL by a clear 4 years. Amazed! As a result ICL offered me the design of the I/O of the 1908A or Chief Engineer of the 1901B, the Basic Language Machine, also known as "Project 52". (Actually it was more complex than that: "Project 52" was a dust-sheet under which ICT/ICL's latest bright idea was placed and changed from time to time, and used as a deice to attract Government subsidies).
There is so much to learn from the latter days at LEO and I hope you will be able to cover some of it later in your project, before it is lost forever.Date : Unknown
Creator : LEO
This exhibit has a reference ID of CH53398. Please quote this reference ID in any communication with the Centre for Computing History.
Click on the Images For Detail