What to do with Transputer Equipment?

In article <20110524222508.f759567b.jkunz at unixag-kl.fh-kl.de>, Jochen Kunz <jkunz at unixag-kl.fh-kl.de> writes: Think MIMD parallel processing (multiple instruction, multiple data) with connectivity determined by the topology of the link network. Since each transputer comes with 4 links, most people did a toroidal mesh. You could add more link chips and make a hypercube, but most people just used what was on-chip without adding additional link fabric interconnect. What you do at that point is up to what algorithms you can efficiently map to the network of CPUs. People used Occam (a variation of Tony Hoare's CSP, Communicating Sequential Processes) or C to program the transputer. Occam has some basic primitives to support reliable parallel processing. I think if you programmed it in C, you could call out to a runtime library that implemented these primitives, or you had to manage it yourself, I can't remember which. -- "The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download <http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/> Legalize Adulthood! <http://legalizeadulthood.wordpress.com>