Optimization of the Load Balancing Policy for Tiled Many-Core Processors

Abstract

Tiled many-core processors (i.e., KNL and the TILE-Gx72 processor), on which processing cores are fitted onto a single chip and cores are interconnected via mesh-based networks, are different from the traditional many-core systems. Their operating system (OS) should be optimized to take into account the unique characteristics (for instance, cores are integrated into a single chip) of tiled many-core processors. This is because these characteristics were not taken into consideration when OSes designed for the traditional multicore (many-core) systems were deployed on tiled many-core processors. In this paper, we propose an optimized load balancing policy to improve the performance of multi-threaded applications. Making a thread select an appropriate idle (lightweight) tile (processing core) across all tiles on the single chip rather than a portion of tiles is able to reduce the overhead triggered by the load balancing policy, the penalty of cache misses because of the scheduling and more threads sharing the same tile (processing core), and the contention for memory controllers due to cache misses. The experimental results demonstrate that the optimized load balancing policy can provide up to 2.7× performance improvement on KNL and mitigate the performance degradation to separate extents on the TILE-Gx72 processor. INDEX TERMS Tiled many-core processors, load balancing, scalability problem.

Automated Power Source Selection System for Uninterrupted Supply: Integration of Main Power, Solar Energy, and Generator Power

“…This method employs a real-time power flow control algorithm to dynamically distribute the load among available power sources. Similarly, the scholar in [10], presented a load balancing and failure recovery technique for multi-power source systems, utilizing a distributed algorithm to balance the load and recover from power source failures [9] [10].…”

Affinity-Based Task Scheduling on Heterogeneous Multicore Systems Using CBS and QBICTM

“…Certain approaches are proposed where the conditions are defined on the basis of which the affinity-based tasks are grouped into a cluster, making sure the grouped data have some kind of dependency [26][27][28][29]. To ensure load balancing in heterogeneous systems, imbalance between scheduling domains NUMA and SMT (simultaneous multithreading) NUMA is identified and load balancing is performed by the balancing interval of scheduling domains [30]. Static and dynamic load balancing techniques are proposed.…”