Synthesis of Parallel Numerical Algorithms

1984 ◽  
pp. 13-43
Author(s):  
Jozef Mikloško ◽  
Vadim Evgenich Kotov
Keyword(s):  
Numerical Algorithms ◽  
Parallel Numerical Algorithms

Parallel Numerical Algorithms

2010 ◽  
pp. 63-64
Author(s):  
Patrick Amestoy ◽  
Daniela di Serafino ◽  
Rob Bisseling ◽  
Enrique S. Quintana-Ortí ◽  
Marian Vajteršic
Keyword(s):  
Numerical Algorithms ◽  
Parallel Numerical Algorithms

PERFORMANCE PREDICTION FOR PARALLEL NUMERICAL ALGORITHMS

1991 ◽  
Vol 03 (01) ◽  
pp. 31-62 ◽  
Author(s):  
K. GALLIVAN ◽  
W. JALBY ◽  
A. MALONY ◽  
H. WIJSHOFF
Keyword(s):  
Performance Prediction ◽  
Numerical Algorithms ◽  
Parallel Numerical Algorithms

Parallel Numerical Algorithms (T. L. Freeman and C. Phillips)

SIAM Review ◽  
1994 ◽  
Vol 36 (4) ◽  
pp. 678-679
Author(s):  
Henk van der Vorst
Keyword(s):  
Numerical Algorithms ◽  
Parallel Numerical Algorithms

Topic 10: Parallel Numerical Algorithms

2012 ◽  
pp. 550-550
Author(s):  
Iain Duff ◽  
Efstratios Gallopoulos ◽  
Daniela di Serafino ◽  
Bora Ucar
Keyword(s):  
Numerical Algorithms ◽  
Parallel Numerical Algorithms

Methods to Overlap Communications in Parallel Numerical Algorithms

1997 ◽  
Vol 08 (02) ◽  
pp. 211-235
Author(s):  
C. Calvin ◽  
L. Colombet ◽  
Ph. Michallon
Keyword(s):  
Execution Time ◽  
Distributed Memory ◽  
Numerical Algorithms ◽  
Total Execution Time ◽  
Distributed Memory Machines ◽  
Matrix Products ◽  
The Matrix ◽  
Intel Paragon ◽  
Matrix Vector ◽  
Parallel Numerical Algorithms

We present in this paper general techniques for overlapping communications in parallel numerical kernels. We describe first some dependencies schemes which can be found in most of numerical parallel algorithms and we apply on these schemes methods based on the change of the granularity of the computational tasks. The choice of the granularity in order to obtain a good overlap depends on the main parameters of the target machines. So we present results of benchmarks executed on two parallel distributed memory machines: a Cray T3D and an Intel Paragon. Then we apply the precedent techniques of overlapping on classical numerical kernels, namely: the matrix-vector and the matrix-matrix products and the mono and bi-dimensional FFT. We have implemented to the overlapped versions of these algorithms on a T3D and a Paragon and tuned the parameters of overlapping in order to minimize the total execution time. The results of these experiments demonstrate the accuracy of this approach.

Sign in / Sign up

Export Citation Format

Share Document