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

Tagged “software performance engineering”