DocumentCode :
3591172
Title :
Cache-conscious scheduling of streaming pipelines on parallel machines with private caches
Author :
Agrawal, Kunal ; Maglalang, Jordyn ; Fineman, Jeremy T.
Author_Institution :
Dept. of Comput. Sci. & Eng., Washington Univ. in St. Louis, St. Louis, MO, USA
fYear :
2014
Firstpage :
1
Lastpage :
12
Abstract :
This paper studies the problem of scheduling a streaming pipeline on a multicore machine with private caches to maximize throughput. The theoretical contribution includes lower and upper bounds in the parallel external-memory model. We show that a simple greedy scheduling strategy is asymptotically optimal with a constant-factor memory augmentation. More specifically, we show that if our strategy has a running time of Q cache misses on a machine with size-M caches, then every “static” scheduling policy must have time at least that of Q(Q) cache misses on a machine with size-M/6 caches. Our experimental study considers the question of whether scheduling based on cache effects is more important than scheduling based on only the number of computation steps. Using synthetic pipelines with a range of parameters, we compare our cache-based partitioning against several other static schedulers that load-balance computation. In most cases, the cache-based partitioning indeed beats the other schedulers, but there are some cases that go the other way. We conclude that considering cache effects is a good idea, but other features of the streaming pipeline are also important.
Keywords :
cache storage; parallel machines; parallel memories; pipeline processing; resource allocation; Ω(Q)-cache; M-cache; asymptotically optimal; cache-based partitioning; cache-conscious scheduling; constant-factor memory augmentation; greedy scheduling strategy; load-balance computation; multicore machine; parallel external-memory model; parallel machine; private cache; static scheduler; streaming pipeline; synthetic pipeline; Analytical models; Computational modeling; Load modeling; Pipelines; Processor scheduling; Schedules; Upper bound;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
High Performance Computing (HiPC), 2014 21st International Conference on
Print_ISBN :
978-1-4799-5975-4
Type :
conf
DOI :
10.1109/HiPC.2014.7116893
Filename :
7116893
Link To Document :
بازگشت