It was felt that Transport Triggered Architecture would provide a good conceptual base for the development of a highly parallelizable FPU architecture. Moving FP data from any virtual register to any TTA functional unit in the FP coprocessor address space takes a single instruction and a single clock cycle.