Redirecting to Thread-level parallelism