Corder: cache-aware reordering for optimizing graph analytics
Proceedings of the 26th ACM SIGPLAN Symposium on Principles and Practice of …, 2021•dl.acm.org
The intrinsic irregular data structure of graphs often causes poor cache utilization thus
deteriorates the performance of graph analytics. Prior works have designed a variety of
graph reordering methods to improve cache efficiency. However, little insight has been
provided into the issue of workload imbalance for multicore systems. In this work, we identify
that a major factor affecting the performance is the unevenly distributed computation load
amongst cores. To cope with this problem, we propose cache-aware reordering (Corder), a …
deteriorates the performance of graph analytics. Prior works have designed a variety of
graph reordering methods to improve cache efficiency. However, little insight has been
provided into the issue of workload imbalance for multicore systems. In this work, we identify
that a major factor affecting the performance is the unevenly distributed computation load
amongst cores. To cope with this problem, we propose cache-aware reordering (Corder), a …
The intrinsic irregular data structure of graphs often causes poor cache utilization thus deteriorates the performance of graph analytics. Prior works have designed a variety of graph reordering methods to improve cache efficiency. However, little insight has been provided into the issue of workload imbalance for multicore systems. In this work, we identify that a major factor affecting the performance is the unevenly distributed computation load amongst cores. To cope with this problem, we propose cache-aware reordering (Corder), a lightweight reordering algorithm that facilitates workload balance as well as cache optimization. Comprehensive performance evaluation of Corder is conducted on various graph applications and datasets. We observe that Corder yields speedup of up to 2.59× (on average 1.47×) over original graphs.