My-Journal, display c. If we were to calculate the sum of all of the data in a LLL of arrays, think about how could parallel processing be used What could be done in parallel?_ What wouldn't work well in parallel?! What data race situations might occur? Since all threads will need to store their answer in a sum total, we would experience a data race if each thread were to write to the same memory at different times; this would result in unpredictable totals! Luckily a thread can have its own private memory (similar to a local variable). As a designer of a parallel algorithm, we need to decide which memory should be shared among threads, which memory should be local for each thread, and where there are critical sections where all but one thread is locked so we can control which thread is accessing a shared variable. For this problem, think about what would be local, shared, and critical: Local per thread Shared among all threads Critical sections Lab #3 Stacks and Queues - Page 94