Jump Over Left Menu
The company most involved with the Atlas Computer Laboratory early on was Ferranti Ltd who had decided to manufacture the Manchester University architecture as a commercial computer called Atlas. Ferranti needed a high profile customer and the Atomic Energy Authority was the most credible in the UK at the time. It needed faith on both sides. On Ferranti's side there had to be a belief in a super-computer market and on the Atomic Energy Authority's side, there had to be a belief that Ferranti could deliver a product at least as good as the rival offering from IBM, the IBM 7030 (Stretch).
Peter Hall of Ferranti and later ICL was the main contact point with the Laboratory over a number of years. Negotiations with ICL over a successor to Atlas also had Peter Hall as a focus point. Arthur Humphreys, the ICL Managing Director at the time, was persuaded that the interim solution to ICL's diverse range of products after the merger with English Electric Marconi was the Ferranti FP6000 that became the ICL 1900 Series. The decision to manufacture a large dual-processor 1908A to maintain ICL's position in the supercomputer market was made while Arthur Humphreys was Managing Director, as was the decision to abandon the project in favour of the ICL New Range that became the ICL 2900 Series. The initial decision by ICL to build the 1908A and later the decision to cancel it had a profound impact on the future of the Atlas Computer Laboratory. If no credible UK supercomputer existed, the Laboratory would most likely have purchased a large scale USA-manufactured machine such as the CDC STAR-100. If the dual-processor 1908A had been delivered with the possibility of an ICL Distributed Array Processor attached, the UK would still have been in the large scale computer market.
Brian Chapman was one of a number of Ferranti personnel who spent the early years of the Atlas Computer laboratory situated at the Laboratory. He arrived quite early on and became one of the Atlas Laboratory team developing the Fortran compiler. He then moved to Aldermaston to carry out a similar function porting the Aldermaston Optimised Fortran compiler to the Atlas II.