Please note, this is archived content and is no longer being updated. Please update your bookmarks and links to point to this permanent URL: http://curation.cs.manchester.ac.uk/computer50/www.computer50.org/kgill/mu5/mu5.html.
Please note, this is archived content and is no longer being updated. Please update your bookmarks and links to point to this permanent URL: http://curation.cs.manchester.ac.uk/computer50/www.computer50.org/kgill/mu5/mu5.html.


Department of Computer Science

MU5



The MU5 Picture Gallery

Planning for a new large computer project started in 1966 with proposals for system architecture and the development of an integrated associative storage technology envisaged as having a speed of 20 time that of the Atlas. These improvements were due to a) improvements in technology b) Overlapping of various activities involved in executing an instruction. i.e. pipelining. c) A more economical and efficient order code. d) Improvements compared with Atlas in operating system performance.

The Main design aim was to produce an architecture capable of running high-level language programs efficiently. This was achived not only by careful formulation of the order code but also by the use of hardware such as associatively accessed name stores to give effect of optimised central registers. A segmented virtual storage with variable sized pages was planned, facillitating time sharing and protection of resources necessary for a real-time multiprogramming enviroment.

A university team of 16 staff and 25 research students worked on the project up until 1971 when the group was increased by 19 engineers from ICL. MU5 started working in mid 1974 when it became possible to run Algol and Fortran programs under operating system control. It was used until 1979 when it was replaced by MU6 a much cheaper computer which produced a similiar performance.

MU5 was the only Manchester computer to recieve direct Government funding when in 1968 the Science Research Council awarded a 5 year grant of just over 630,000 pounds to the department. At the same time ICL, who had taken over Ferranti's computer division, agreed to give assistance in terms of production facilties and manpower. Part of the SRC grant was used to purchase an ICL 1905E computer which was used between early 1969 and early 1972 for detailed MU5 logic simulation and as a test bed for an associative store.

MU5 was the only Manchester computer which did not have a direct commercial counterpart. Although many architectural concepts of the ICL 2900 series were derived from those of MU5.


Machine hall | Main Hall

Copyright The University of Manchester 1996