colin@inmos.UUCP (Colin Whitby-Strevens) (07/22/88)
Don't keep the knowledge of the transputer to yourself! Has anyone seen the article in June 88 issue of IEEE Computer by Rayfield and Silverman on the Armstrong Multiprocessor (Brown multiprocessor). They claim reconfigurability (or re-pluggability) and extensive support for message passing as being unique features. Their heart appears to be in the right place, but they are suffering enormously from their chosen hardware solution, and getting unnecessarily negative results. They seem to be totally unaware of the transputer based machines. If anyone knows these guys, would they like to interact with them, and introduce them (in the nicest possible way) to observations such as that linear scalability has been demonstrated for a wide range of applications for systems containing 100's of transputers, that the performance is two orders of magnitude faster, that the engineering is orders of magnitude simpler, and that there is a proper programming model. Possibly some correspondence in Computer open channel would also be a good idea.