Paper Info
Title
Anatomy of GPU Memory System for Multi-Application Execution
Abstract
As GPUs make headway in the computing landscape spanning mobile platforms, supercomputers, cloud and virtual desktop platforms, supporting concurrent execution of multiple applications in GPUs becomes essential for unlocking their full potential. However, unlike CPUs, multi-application execution in GPUs is little explored. In this paper, we study the memory system of GPUs in a concurrently executing multi-application environment. We first present an analytical performance model for many-threaded architectures and show that the common use of misses-per-kilo-instruction (MPKI) as a proxy for performance is not accurate without considering the bandwidth usage of applications. We characterize the memory interference of applications and discuss the limitations of existing memory schedulers in mitigating this interference. We extend the analytical model to multiple applications and identify the key metrics to control various performance metrics. We conduct extensive simulations using an enhanced version of GPGPU-Sim targeted for concurrently executing multiple applications, and show that memory scheduling decisions based on MPKI and bandwidth information are more effective in enhancing throughput compared to the traditional FR-FCFS and the recently proposed RR FR-FCFS policies.
Year
DOI
Venue
2015
10.1145/2818950.2818979
MEMSYS
Field
DocType
Citations 
Headway,Computer architecture,Computer science,Interference theory,Bandwidth (signal processing),Interference (wave propagation),Virtual desktop,Throughput,Cloud computing,Embedded system,CUDA Pinned memory
Conference
26
PageRank 
References 
Authors
0.68
32
9
Name
Order
Citations
PageRank
Adwait Jog156823.32
Onur Kayıran235613.47
Tuba Kesten3260.68
Ashutosh Pattnaik41134.70
Evgeny Bolotin557925.80
Niladrish Chatterjee626711.53
Stephen W. Keckler73404201.71
Mahmut T. Kandemir87371568.54
Chita R. Das9103859.34