Roy Pedlar
Photoshoot at Bracknell for ICL 2970 launch in October 1974.
Although probably the most widely seen photograph of an ICL 2900, this system was not really a standard production model but had various non-standard devices attached for it's role
as a conversion platform for customers wishing to move from the ICL 1900 and ICL System 4
ranges to ICL 2900. This system's role was superceded when the ICL 2960 was introduced with Direct Machine Emulation (DME), as well as the native Virtual Machine Envionment (VME), around 3 years later. This delay period is criticised in the book "ICL - A Business and Technical History " (1989) by Martin Campbell-Kelly on page 326. The delay was probably necessary to prolong the sales of the ICL 1903T and ICL 1904S, which also had common peripherals with the ICL 2970 (MT4, LP1500,CR1200) and solid state stores, but the expectation of those of us who came from an IBM/RCA system architecture background was that direct emulation would be provided on the ICL 2970 in 1974 to provide a more powerful addition/replacement for the ICL System 4-72. In particular this was as the 4-72 was a good transaction processing machine for various important, high profile customers who needed more MIPS - but this was not to be. It was not clear in 1974 whether the IBM order code microcode for the ICL 2970 existed or was just not made available. The proposed ICL 2970 Mark 2 was cancelled, leaving the blanking plates for the LED displays by the side of the speaker (which would have presumably have been moved elsewhere as happened to it on the ICL2966) as can be seen on my photos of the ICL 2970 Engineers Panel (see the photo to the immediate right hand side of this photo), and what later became the ICL 2972 was not the Mark 2 P3+ but a slowed down version of the P4 OCP grafted into the ICL 2970 central units instead. However this made the ICL 2972 a hard-wired processor system that could only run VME.
Photoshoot at Bracknell for ICL 2970 launch in October 1974.
Although probably the most widely seen photograph of an ICL 2900, this system was not really a standard production model but had various non-standard devices attached for it's role
as a conversion platform for customers wishing to move from the ICL 1900 and ICL System 4
ranges to ICL 2900. This system's role was superceded when the ICL 2960 was introduced with Direct Machine Emulation (DME), as well as the native Virtual Machine Envionment (VME), around 3 years later. This delay period is criticised in the book "ICL - A Business and Technical History " (1989) by Martin Campbell-Kelly on page 326. The delay was probably necessary to prolong the sales of the ICL 1903T and ICL 1904S, which also had common peripherals with the ICL 2970 (MT4, LP1500,CR1200) and solid state stores, but the expectation of those of us who came from an IBM/RCA system architecture background was that direct emulation would be provided on the ICL 2970 in 1974 to provide a more powerful addition/replacement for the ICL System 4-72. In particular this was as the 4-72 was a good transaction processing machine for various important, high profile customers who needed more MIPS - but this was not to be. It was not clear in 1974 whether the IBM order code microcode for the ICL 2970 existed or was just not made available. The proposed ICL 2970 Mark 2 was cancelled, leaving the blanking plates for the LED displays by the side of the speaker (which would have presumably have been moved elsewhere as happened to it on the ICL2966) as can be seen on my photos of the ICL 2970 Engineers Panel (see the photo to the immediate right hand side of this photo), and what later became the ICL 2972 was not the Mark 2 P3+ but a slowed down version of the P4 OCP grafted into the ICL 2970 central units instead. However this made the ICL 2972 a hard-wired processor system that could only run VME.