Parallel computation of electromagnetic fields on transputers

About ten years ago, after a long struggle with our bureaucracy, I got a Tektronix 4051, with BASIC interpreter and 16 kB (kilobytes) of memory. I felt really happy, because I could edit programs and input data locally, even when all ports of the CDC 6500 were busy, or when the mainframe was not runnin for one of numerous reasons. But it still was a kght against the system, rather than working with a computer: Hours of data transfers and colleagues working on the same machine forced me to interrupt my investigations quite often