A number of meetings took place between Peter Lever and Roger Vinnicombe of ICL's Government Sector and Rob Witty and myself during October.
Peter Lever had talked to ICL Management and had received permission to approach Three Rivers. At the same time, we had contacted Three Rivers to see if they would be sympathetic to an approach from ICL. Three Rivers did need further equity and were interested in doing some kind of a deal. As a result, a meeting was held with ICL to discuss possible strategies on 10 October 1979. It was clear that nothing would happen in ICL unless its senior management were behind it. The relevant person to convince initially was Ninian Eadie, head of ICL's Government Sector. We saw two options:
It was agreed that Roger Vinnicombe would visit Three Rivers on 16 October and RAL would write to Ninian Eadie urging that some positive action be taken by ICL.
The text of the letter from F R A Hopgood to Ninian Eadie on 18 October 1979 is given below.
As you are probably aware, the SRC has a programme of research in University departments into Distributed Computing Systems which is coordinated from the Rutherford Laboratory. One of the aims is to foster the transfer of research developments between SRC and the relevant sectors of British industry. In particular, there is potentially a large market developing out of research into man-machine interaction and distributed computing systems. We were pleased to hear of ICL's research programme in this area and, from Peter Lever, I understand that the company is keen to collaborate with SRC and the DCS academic research community.
The SRC's sponsorship of distributed computing research has 'priority' status and is organised as the coordinated programme of research into Distributed Computing Systems (DCS). Some 40 academic research projects are currently active and the SRC expects to invest £4-5 million in this area between 1977 and 1984. A primary objective of the DCS Programme is to actively link university research and British industry. The DCS research is relevant to the large and developing market place which ranges from the new generation of 'mini' systems, office automation, the home/hobby market, the research and education (especially schools) markets to process control and data capture.
It is clear from experience in the USA that the collaborative synthesis of academic and industrial expertise is the key to the rapid development of successful new products.
The Rutherford Laboratory and the DCS programme need a powerful highly interactive single-user computing system centred around a high quality display (at least capable of displaying an A4 page) similar to the Xerox PARC Alto which has set the pace in this field. A system borrowing greatly from the Alto experience is the Three Rivers Company's PERQ which has just appeared on the market. Unfortunately, there is no British competitor. I feel that ICL could benefit significantly from tackling this gap in the market and you might like to consider two possible strategies should ICL wish to exploit the current situation.
Of these two strategies I would lean in favour of the first. ICL West Gorton are very well placed to undertake the development of a superior PERQ thanks to the work of C Portman whose Project Little has laid the foundations of solutions to problems such as system reliability, video data compression and data security which the present PERQ type systems cannot begin to tackle but which are vital for the wider market. ICL's expertise naturally complements that of Three Rivers in the design of high performance display systems.
The first strategy gives ICL the opportunity to establish themselves in this market quickly whilst allowing West Gorton to put their full effort into what could be a major British force in the world market.
I feel that this is an important opportunity which should not be missed and would give ICL a prominent position in this market.
Dr Manning included a covering note stressing that RAL considered this area of great importance to the SRC and believed that it is an area of considerable National importance. No British Company is at the moment in the running and I believe that it is an area of great growth and one in which a British Company should be able to make a rapid impact with a relatively small investment.
ICL were impressed by the PERQ hardware although the software appeared to be in a rudimentary state and it was estimated that a viable software environment would not be available until March 1980. ICL had arranged with Three Rivers to send somebody from their manufacturing facility at Utica to Three Rivers to assess whether there were any problems in mass production of the system.
ICL had taken an option on a PERQ to be delivered in December 1979 to have a system to show around in order to test the size of the market. In their cautious way, ICL asked if SERC would be interested in purchasing the system from them if they did not go ahead. As we were interested in several systems, we agreed to do that. The DCS requirement was independent of whether ICL were involved or not. The only condition that was placed was that it had to be delivered to RAL within 6 months and we would pay 75% of the purchase price.
Meanwhile, Ninian Eadie had replied positively to the letters from RAL on 25 October. He confirmed that ICL believed this to be an important growth area and he welcomed any collaboration with SRC that could be established. Neil Davenport (now of CRAY-UK) who was responsible for the Education and Science Region of the Government Sector was designated as the person responsible for pushing the project forward.
A meeting was held between RAL Staff and ICL on 15 November to discuss a sensible strategy to put forward to both sets of management. The major conclusions of the meeting were:
The management meeting between RAL and ICL, also on 15 November, confirmed the above strategy. ICL agreed to buy the PERQ with the condition that they could resell to RAL. The Company wanted some commitment from SERC as to the level of purchasing that SERC would do if ICL moved in this direction. ICL wanted advance orders of £400K from SERC with ICL committing to deliver in August 1980. This was clearly something that RAL was not prepared to do at this stage. We had not had delivery of a system for evaluation and the funding within SERC was such that only Council would be in a position to make such a decision.
At this meeting, the possible involvement of ERCC was raised. One of the ICL Directors had mentioned it to Dr Thomas of ERCC and it was agreed to hold a tripartite meeting in December.
On 6 December 1979, the Engineering Board received a Grant Application from Newcastle requesting two PERQ systems to explore the potential of using the PERQ computer as a sophisticated bitmap display processor as a front end to a UNIX system.
Three Rivers had quoted a production price to Newcastle which, with VAT and import duty, came to £29,220, at the current exchange rate, for each system.
The visit by ICL Utica staff to Three Rivers was not a sparkling success. The Utica person clearly was unimpressed by the size of the company and its method of production. However, his report indicated that the PERQ was a very powerful modern graphics control workstation.
He saw possible environmental problems. There were no provisions for static discharge immunity or VDE approval. He believed changes would have to be made to the system to achieve that. He was able to cost the components in the system and give ICL an estimate of what it would cost ICL to manufacture in bulk. We never received details of this costing but later meetings suggested that the Three Rivers price was viable as long as the product was sold in bulk.
The report was on the whole positive. However, considerable damage was caused to relationships between ICL and Three Rivers (particularly Brian Rosen, the designer of the PERQ).
Between October and December 1979, sufficient work had been done that an agreement between ICL and Three Rivers looked possible.
With ICL's financial state at that time, they were reluctant to proceed without a significant level of commitment from SERC which would not be possible without getting Council approval.
Many more people were becoming aware of the PERQ and SRC had started receiving grant applications requesting purchases from Three Rivers.
Meanwhile, delivery of the RAL system had slipped from November to January due to some hardware faults on the production system that had taken longer than expected to diagnose and cure.
The major tasks in the first quarter of 1980 were to assess the system, and make the relevant Committees in SERC aware of the product and its potential. Finally, for ICL to finalise the deal with Three Rivers, it would be necessary for Council to make some commitment to purchase a significant number.
