Tags

instructors get_involved Chip multiprocessor atomic cilk cilk_for cilk_scope cilk_spawn cilk_sync commutative operation cpu span core critical section data race determincacy race distributed memory determinism deadlock global variable execution time fake lock hyperobject instruction linear speedup Cilksan lock contention lock multicore nondeterminism mutex nonlocal variable parallel loop multiprocessor opencilk false sharing perfect linear speedup parallelism process processor race condition reducer response time receiver running time sequential consistency serial execution scale down scale up serial semantics serial projection concurrent agent software performance engineering spawn speedup strand task-parallel sync throughput thread work worker view work stealing events competition knot shared memory