TY - JOUR
T1 - Empirical study of dynamic scheduling on rings of processors
AU - Barrows, Miranda E.
AU - Gregory, Dawn E.
AU - Gao, Lixin
AU - Rosenberg, Arnold L.
AU - Cohen, Paul R.
N1 - Funding Information:
The authors wish to thank Nadia Perez of the University of Catania for programming assistance. The work of Miranda Barrows and Arnold Rosenberg was supported in part by NSF Grants CCR-92-21785 and CCR-97-10367. The work of Dawn Gregory was supported in part by a NSF Graduate Fellowship Award and by DARPA/RL Contract F30602-93-C-0100. The work of Lixin Gao was supported in part by NSF Grant NCR-9729084.
PY - 1999/9
Y1 - 1999/9
N2 - We empirically analyze and compare two low-overhead, deterministic policies for scheduling dynamically evolving tree-structured computations on rings of identical processing elements (PEs). Our computations have each task either halt or spawn two independent children and then halt; they abstract, for instance, computations generated by multigrid methods. Our simpler policy, KOSO, has each PE keep one child of a spawning task and pass the other to its clockwise neighbor in the ring; our more sophisticated policy, KOSO*, operates similarly, but allows child-passing only from a more heavily loaded PE to a more lightly loaded one. Both policies execute waiting tasks in increasing order of their depths in the evolving task-tree. Our study focuses on two conjectures: (a) Both policies yield good parallel speedup on large classes of the computations we study. (b) Policy KOSO* outperforms policy KOSO in many important situations. We verify these conjectures via a suite of experiments, supplemented by supporting mathematical analyses. We view our methodology of experimental design and analysis as a major component of our study's contribution, which will prove useful in other such studies.
AB - We empirically analyze and compare two low-overhead, deterministic policies for scheduling dynamically evolving tree-structured computations on rings of identical processing elements (PEs). Our computations have each task either halt or spawn two independent children and then halt; they abstract, for instance, computations generated by multigrid methods. Our simpler policy, KOSO, has each PE keep one child of a spawning task and pass the other to its clockwise neighbor in the ring; our more sophisticated policy, KOSO*, operates similarly, but allows child-passing only from a more heavily loaded PE to a more lightly loaded one. Both policies execute waiting tasks in increasing order of their depths in the evolving task-tree. Our study focuses on two conjectures: (a) Both policies yield good parallel speedup on large classes of the computations we study. (b) Policy KOSO* outperforms policy KOSO in many important situations. We verify these conjectures via a suite of experiments, supplemented by supporting mathematical analyses. We view our methodology of experimental design and analysis as a major component of our study's contribution, which will prove useful in other such studies.
UR - http://www.scopus.com/inward/record.url?scp=0032645114&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0032645114&partnerID=8YFLogxK
U2 - 10.1016/S0167-8191(99)00039-3
DO - 10.1016/S0167-8191(99)00039-3
M3 - Article
AN - SCOPUS:0032645114
SN - 0167-8191
VL - 25
SP - 1063
EP - 1079
JO - Parallel Computing
JF - Parallel Computing
IS - 9
ER -