high performance linpack

12
LINPACK Introduction CSTL HPC IT Mu Wei

Upload: wei-mu

Post on 08-Jul-2015

138 views

Category:

Technology


0 download

DESCRIPTION

HPL is a software package that solves a (random) dense linear system in double precision (64 bits) arithmetic on distributed-memory computers. It can thus be regarded as a portable as well as freely available implementation of the High Performance Computing Linpack Benchmark. The algorithm used by HPL can be summarized by the following keywords: Two-dimensional block-cyclic data distribution - Right-looking variant of the LU factorization with row partial pivoting featuring multiple look-ahead depths - Recursive panel factorization with pivot search and column broadcast combined - Various virtual panel broadcast topologies - bandwidth reducing swap-broadcast algorithm - backward substitution with look-ahead of depth 1. The HPL package provides a testing and timing program to quantify the accuracy of the obtained solution as well as the time it took to compute it. The best performance achievable by this software on your system depends on a large variety of factors. Nonetheless, with some restrictive assumptions on the interconnection network, the algorithm described here and its attached implementation are scalable in the sense that their parallel efficiency is maintained constant with respect to the per processor memory usage.

TRANSCRIPT

Page 1: High performance LINPACK

LINPACK Introduction

CSTL HPC IT Mu Wei

Page 2: High performance LINPACK

Agenda

• LINPACK

• LINPACK & MATRIX

• HPL TEST

Page 3: High performance LINPACK

What’s LINPACK

• A collection of Fortran subroutines that analyze and sovlve linear equations and linea least-squares problems

• LINPACK100,LINPACK1000,HPL

Page 4: High performance LINPACK

LINPACK AND MATRIX

• Based on decompositional approach to numerical linear algebra

• Divide computational problem into two parts

• Ax=b

Page 5: High performance LINPACK

Organized around matrix decompositions

LINPACK AND MATRIX

LU Cholesky

QR singular value

Page 6: High performance LINPACK

NAME OF LINPACK

• T - Type of arithmetic

• XX - Reflect a fundamental division

• YY - Specifies task the subroutine is to perform

Name of LINPACK subroutine is divided into a prefix, an infix, a suffix as follow

Page 7: High performance LINPACK

EFFICIENCY

• Column orientation of the algorithms

• Use of Basic Liear Algebra Subprograms

Effects of two aspects of LINPACK on efficiency

Page 8: High performance LINPACK

How to Setup HPL

• Download Open source HPL

• Prequisites: 1) BLAS; 2)MPI

• Edit Data File

• Compile: $ make arch=Linux_PII_CBLAS

• Execute

Page 9: High performance LINPACK

DATA FILE$ cat HPL.dat HPLinpack benchmark input file Innovative Computing Laboratory, University of Tennessee HPL.out output file name (if any) 6 device out (6=stdout,7=stderr,file) 4 # of problems sizes (N) 29 30 34 35 Ns 4 # of NBs 1 2 3 4 NBs 0 PMAP process mapping (0=Row-,1=Column-major) 3 # of process grids (P x Q) 2 1 4 Ps 2 4 1 Qs

Page 10: High performance LINPACK

Key Parameters

• Estimate the LARGEST PROBLEM SIZE

• Expected Execution Time

• Expected Gflops

• Sample - Choose N=14k, then Time=52s

Page 11: High performance LINPACK

RESULT

Page 12: High performance LINPACK

Q & A