Guided self-scheduling: A practical scheduling scheme for parallel supercomputers
IEEE Transactions on Computers
Proceedings of the 1988 ACM/IEEE conference on Supercomputing
ICS '90 Proceedings of the 4th international conference on Supercomputing
Hi-index | 0.00 |
Vectorization reduces CPU time, whereas parallelization leads to a reduction of wall-clock time at the cost of more CPU cycles. Application programs will in general not be perfectly parallel. Therefore a static scheduling scheme for processor usage will lead to a waste of cycles. In contrast with this, dynamic allocation of processors will prevent CPUs from being idle as long as there is work to do.