Mannerisms

Actually Bristol had IC design engineers before Inmos came along - Fairchild had a design team in Bristol in the 1970's; I was one of them.

Keith Sabine is quite right. Actually, the presence of the IC design industry in Bristol is due to one Sally, a microbiology PhD at Bristol University in the early 1970's. This is how it happened. Sally was the former student sweetheart of Tony Bell, he designed TIs first one chip calculator in the USA as an ex-pat Englishman. In 1970/1 in Silicon valley the IC design industry was growing very fast and Fairchild had more problems than most in retaining engineers. Everyone wanted to start at Fairchild but get better salaries elsewhere as soon as they were trained. Fairchild decided to set up design centres away from other semiconductor companies as a strategy to retain engineers. They recruited Tony Bell from TI to set up a design centre in Europe - anywhere he liked as long as there was no other semiconductor industry nearby. Tony Bell set up a design centre, initially four people, in an office over a fish & chip shop in Gloucester Rd. He chose Bristol because Sally had come here and he hoped to re-start their relationship (which he did, though I believe it was quite stormy). The first product they designed in Bristol was a custom VLSI one chip printing calculator for Olivetti in 4 phase depletion load PMOS. Fairchild Bristol had grown to about 30 people by the time I joined in 1975 and mostly designed 4000 series CMOS commodity logic. They were in the top floor (6th) of the Natwest Court, just off the Bristol centre. In 1978 we switched to FAST (oxide isolated shottky TTL) design and Tony Bell returned to CalTech where Fairchild financed a PhD in automating layout. The new manager was Peter Caville, one of the original 4 from 1972 Bristol Fairchild office. When Inmos was announced Peter was approached by Ian Barron and offered the job of Design Manager there, being the manager of the biggest IC design centre in Britain at the time. Peter agreed to take the job as long as he didn't have to move and this is the reason that Inmos came to Bristol. The original Inmos office in Whitefriars was less than 100 yards from the Fairchild office and engineers from both offices often drank together on Friday lunchtimes. The rest is the history you know. I worked with Peter at Inmos (I headed up the team of four doing the hardware design of the original Transputer CPUs), as did many other former Fairchild Bristol employees. Fairchild was bought by Schlumberger and in 1980 they moved the design centre to Reading to prevent Inmos pinching any more employees. Actually very few Fairchild Bristol employees, 2 or 3, moved to Reading despite generous relocation terms, and Fairchild had to start more or less from scratch in Reading. I contracted for them there for about 6 months in 1982 where my main task was educational, training recent grads to design in HCMOS. In 1983/4 I worked for Fairchild again in South Portland Maine USA introducing the FACT family of advanced CMOS logic. Personally, I regard the Transputer as a failure (I was offered a contract to fix the T9000 in 1994 but refused it) and David May as the priciple architect of the failure, despite the plaudits he got from high places. Inmos did not fufil the hype of 'a completely new processing paradigm' and many good ideas were missed in the battles of egos that went on at Inmos. The company was hamstrung from the start by having hired mostly software people and very few hardware people because the Transputer was going to be 'designed by software'. (Not the way to get a competitive product, even now).

Thank you Chez that's a fascinating story. Obviously Sally deserves an honorary doctorate too, best wishes, David.