TY - GEN
T1 - Out-of-core and pipeline techniques for wavefront algorithms
AU - Caron, Eddy
AU - Desprez, Frédéric
AU - Suter, Frédéric
PY - 2005
Y1 - 2005
N2 - Several numerical computation algorithms exhibit dependences that lead to a wavefront of the computation. Depending on the data distribution chosen, pipelining communication and computation can be the only way to avoid a sequential execution of the parallel code. The computation grain has to be wisely chosen to obtain at the same time a maximum parallelism and a small communication overhead. On the other hand, when the size of data exceeds the memory capacity of the target platform, data have to be stored on disk. The concept of out-of-core computation aims at minimizing the impact of the I/O needed to compute on such data. It has been applied successfully on several linear algebra applications. In this paper we apply out-of-core techniques to wavefront algorithms. The originality of our approach is to overlap computation, communication, and I/O. An original strategy is proposed using several memory blocks accessed in a cyclic way. The resulting pipeline algorithm achieves a saturation of the disk resource which is the bottleneck in out-of-core algorithms.
AB - Several numerical computation algorithms exhibit dependences that lead to a wavefront of the computation. Depending on the data distribution chosen, pipelining communication and computation can be the only way to avoid a sequential execution of the parallel code. The computation grain has to be wisely chosen to obtain at the same time a maximum parallelism and a small communication overhead. On the other hand, when the size of data exceeds the memory capacity of the target platform, data have to be stored on disk. The concept of out-of-core computation aims at minimizing the impact of the I/O needed to compute on such data. It has been applied successfully on several linear algebra applications. In this paper we apply out-of-core techniques to wavefront algorithms. The originality of our approach is to overlap computation, communication, and I/O. An original strategy is proposed using several memory blocks accessed in a cyclic way. The resulting pipeline algorithm achieves a saturation of the disk resource which is the bottleneck in out-of-core algorithms.
UR - http://www.scopus.com/inward/record.url?scp=33746315909&partnerID=8YFLogxK
U2 - 10.1109/IPDPS.2005.318
DO - 10.1109/IPDPS.2005.318
M3 - Conference contribution
AN - SCOPUS:33746315909
SN - 0769523129
SN - 0769523129
SN - 9780769523125
T3 - Proceedings - 19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005
SP - 9b
BT - Proceedings - 19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005
T2 - 19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005
Y2 - 4 April 2005 through 8 April 2005
ER -