Oral history of British science
Edwards, David (Dai) (Part 6 of 13). An Oral History of British Science.
The British Library Board acknowledges the intellectual property rights of those named as contributors to this recording and the rights of those not identified.
Legal and ethical usage »
Computer Software; Electronics
Interviewee's home, Preston
Edwards, David B.G., 1928- (speaker, male)
Lean, Thomas (speaker, male)
Part 6: Mentions completing PhD in 1954. Involvement with move of the Mark 1 and Meg computers to a new building on Dover Street, for the computing service and electronics development work, subsequent replacement of Meg by Mercury in 1958 and scrapping of Mark 1. Remarks on thinking about the future, circa 1955, Mercury user feedback on floating point and Professor Flower's remarks on Mercury. [4:30] Detailed technical description of Atlas computer: one level store, the predecessor of virtual memory; fixed store, a random access read only memory for system commands; expanded B-line; red tape instructions; floating point accumulator; improved multiplier; control registers; interrupts and peripherals; tape decks [13:50]Further comments on use of Atlas: communications to Jodrell Bank, Nottingham and Edinburgh Universities, with reference to problems; software and the supervisor program, with reference to work of Tony Brooker and David Howerth; Atlas working practices, computer operators, isolation of users from machine and costings. [20:15] Discussion on time-sharing peripherals, such as x-ray goniometer, rather than user dumb terminals, multiprogramming. Remarks on inauguration of Atlas by Sir John Cockcroft in 1962 and start of computer service in 1963, but with Atlas specific software still in development. Further remarks on inauguration ceremony. [23:55] Remarks on accommodation of Atlas. Comments on early development of Atlas c.1958, MUSE, target of a microsecond per instruction. [25:55] Comments on a report on Atlas from the NRDC's working party on a British national computer: speed, random access memory, peripherals, meetings with Tom Kilburn, small team size, time deadlines, funding, Ferranti, possible cooperation with Royal Radar Establishment at Malvern, project leadership, attitudes of committee. [33:20] Remarks on start of Ferranti invlovement c.1959, with anecdote about meeting Ianto Warburton. Further remarks on NRDC: early patent meetings with F.C. Williams and vice chancellor; expert committee, of which Tom Kilburn was a member. Remarks on F.C. Williams' move away from computing, difference of opinion on hardware floating point. Remarks on Manchester computing team size, overseas research students, including a Czechoslovakian refugee, and work of research students. [39:20] Remarks on Atlas reliability concerns, experience with small transistor computer, point contact transistors and junction transistors. Comments on faster logic circuits, surface barrier transistors, SB240 and 2M501 transistors, drift or graded base transistor, their applications in Atlas, necessity of using cheaper transistors. [44:10] Remarks on parallel addition and the use of surface barrier transistors in high speed operations and patenting of the fast carry technique in late 1950s. Further remarks on choice of cheaper transistors, power considerations, use of coaxial cables to transmit signals. [49:50] Remarks on quest for speed in Atlas: parallel operation, faster circuits, discussion with component manufacturers, smaller and faster memory cores, transistors [53:05] Comments on running partially fluxed memory cores. Remarks on beginning of Ferranti involvement, change of name to Atlas. Short anecdote about almost naming the machine BISON - Built In Spite Of NRDC. Further comments on support from Ferranti, sub-contraction to Plessey, who made surface barrier transistors on licence from Philco, manufacture of Atlas memory, high speed interfaces. [58:45] Further remarks on experiments on core storage and working with industry. Remarks on working with printed circuit boards, initially supplied by Ferranti, later made in department. Description of partial flux experiments on core storage. Short story about accident that destroyed a rack of expensive transistors. [1:04:20] Further remarks on hazards. Remarks on time pressures, construction of adders, competition with other groups. [1:06:15] Comments on feedback from users, user transition from Mark 1 to Mercury. Comments on Atlas user service, computer service staff, later change to terminals on other machines. [1:09:45] Remarks on changes in computer service staff skills, using Atlas, engage and disengage buttons and interrupts. Remarks on computer users, computer supervisor. [1:13:40] Discussion of interaction with software developers during the Atlas development, including Tony Brooker, David Howerth, Derek Morris [DM] and Bruce Paine.
Interview with computer scientist and electrical engineer David (Dai) Edwards.