مشاهده مشخصات مقاله
Implementing new algorithms in MPICH for all the collective operations
Authors |
-
Reza Mirani
-
Mohammad Reza Heidari
|
Conference |
دوازدهمین کنفرانس بینالمللی سالانه انجمن کامپیوتر ایران |
Abstract |
We report on our work on improving the performance of
collective operations in MPICH on clusters connected by
switched networks.
For each collective operation, we use multiple algorithms
depending on the message size, with the goal of
minimizing latency for short messages and minimizing
bandwidth usage for long messages. Although we have
implemented new algorithms for all MPI collective
operations, because of limited space we describe only the
algorithms for allgather, broadcast,reduce-scatter, and
reduce. We present performance results using the
SKaMPI benchmark on a Myrinet-connected Linux
cluster and an IBM . In all cases, the new algorithms
significantly outperform the old algorithms used in
MPICH on the Myrinet cluster, and, in many cases, they
outperform the algorithms used in IBM’s MPI . |
قیمت |
-
برای اعضای سایت : 100,000 Rial
-
برای دانشجویان عضو انجمن : 20,000 Rial
-
برای اعضای عادی انجمن : 40,000 Rial
|
خرید مقاله
|
|