Dynamic Co-scheduling Driven by Main Memory Bandwidth Utilization

被引：16

作者：

Breitbart, Jens ^{[1
]}

Pickartz, Simon ^{[2
]}

Lankes, Stefan ^{[2
]}

Monti, Antonello ^{[2
]}

Weidendorfer, Josef ^{[3
]}

机构：

[1] Bosch Chassis Syst Control, Stuttgart, Germany

[2] Rhein Westfal TH Aachen, Inst Automat Complex Power Syst, Aachen, Germany

[3] Tech Univ Munich, Chair Comp Architecture, Dept Informat, Munich, Germany

来源：

2017 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER) | 2017年

关键词：

D O I：

10.1109/CLUSTER.2017.59

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Most applications running on supercomputers achieve only a fraction of a system's peak performance. It has been demonstrated that the co-scheduling of applications can improve the overall system utilization. However, following this approach, applications need to fulfill certain criteria such that the mutual slowdown is kept at a minimum. In this paper, we present an HPC scheduler that applies co-scheduling and utilizes virtual machine migration for a re-orchestration of applications at runtime based on their main memory bandwidth requirements. Given a job queue consisting of main memory-bound applications and compute-bound applications, we can see a throughput increase of up to 35% while at the same time reducing energy consumption by around 30%.

引用

页码：400 / 409

页数：10