In the context of single processor design, the two main methodologies used to accomplish TLP are chip-level multiprocessing (CMP) and simultaneous multithreading (SMT).