An Efficient Technique of Application Mapping and Scheduling on Real-Time Multiprocessor Systems for Throughput Optimization

被引:4
作者
Liu, Weichen [1 ,2 ]
Xiao, Chunhua [2 ]
机构
[1] Chongqing Univ, Key Lab Dependable Serv Comp Cyber Phys Soc, Minist Educ, Chongqing, Peoples R China
[2] 174 Shazheng St, Chongqing 400044, Peoples R China
关键词
Performance; Algorithms; Multiprocessor; scheduling; optimization; satisfiability; DATA-FLOW PROGRAMS; ALLOCATION; GRAPHS;
D O I
10.1145/2950051
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Multiprocessor systems are becoming ubiquitous in today's embedded systems design. In this article, we address the problem of mapping an application represented by a Homogeneous Synchronous Dataflow (HSDF) graph onto a real-time multiprocessor platform with the objective of maximizing total throughput. We propose that the optimal solution to the problem is composed of three components: actor-to-processor mapping, retiming, and actor ordering on each processor. The entire problem is systematically modeled into a Boolean Satisfiability (SAT) problem such that the optimal solution can be guaranteed theoretically. In order to explore the vast solution space more efficiently, we develop a specific HSDF theory solver based on the special characteristics of the timed HSDF, and integrate it into the general search framework of the SAT solver. Two alternative integration methods based on branch-and-bound are presented to achieve early branch pruning in the search space; thus, the scalability is greatly improved. Extensive performance evaluation on synthetic examples and a case study on the realistic H.264 Video Decoder show that our approach provides as much as 76.9% throughput improvement, and is scalable to industry-sized applications.
引用
收藏
页数:25
相关论文
共 32 条
[1]  
Alvarez Mauricio, 2009, P 4 COL COMP C 4CCC
[2]  
Bonfietti A, 2010, DES AUT TEST EUROPE, P897
[3]  
Cong J, 2007, FPGA 2007: FIFTEENTH ACM/SIGDA INTERNATIONAL SYMPOSIUM ON FIELD-PROGRAMMABLE GATE ARRAYS, P99
[4]   Faster maximum and minimum mean cycle algorithms for system-performance analysis [J].
Dasdan, A ;
Gupta, RK .
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, 1998, 17 (10) :889-899
[5]   Exhaustive scheduling and retiming of digital signal processing systems [J].
Denk, TC ;
Parhi, KK .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-ANALOG AND DIGITAL SIGNAL PROCESSING, 1998, 45 (07) :821-838
[6]   An extensible SAT-solver [J].
Eén, N ;
Sörensson, N .
THEORY AND APPLICATIONS OF SATISFIABILITY TESTING, 2004, 2919 :502-518
[7]  
He W.S., 2014, ABSTR APPL AN, V2014, P1, DOI [DOI 10.1186/S13568-014-0054-7, DOI 10.1155/2014/682015]
[8]  
Heras F, 2008, J ARTIF INTELL RES, V31, P1
[9]   STATIC SCHEDULING OF SYNCHRONOUS DATA FLOW PROGRAMS FOR DIGITAL SIGNAL-PROCESSING [J].
LEE, EA ;
MESSERSCHMITT, DG .
IEEE TRANSACTIONS ON COMPUTERS, 1987, 36 (01) :24-35
[10]  
Liu WC, 2015, INT CONF COMPIL ARCH, P127, DOI 10.1109/CASES.2015.7324553