Fast and parallel video encoding by workload balancing

The issue of balancing the macroblocks (MB) computing workload across the processors are explored. These includes, the prediction of the workload based on the previous frame workload and the scheduling of the MB bounded by the locality constraint. The algorithm was implemented on an IBM SP2, and the results showed that the reduction in the worst case delay is around 19-23%, with both the prediction and scheduling overhead taken into account. Because of the critical path reduction, the overall processor utilization was increased and the overall coding rate improved.published_or_final_versio