Tags

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