Tags

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