Inmos Transputer Architecture
Execution path
SLTransputerCPU -> instruction fetch -> decode -> operand/workspace evaluation -> integer/FPU execution -> process scheduler -> channel/link interface -> memory -> completion
Concurrency
Unlike conventional single-threaded CPU models, the Transputer architecture explicitly supports process scheduling and communication channels. SLeeLa therefore models scheduler, channel, and link resources as first-class architectural components.
Profiles
T4xx and T8xx generations differ in memory, links, floating point, and implementation details. Those differences remain profile-specific.