Statistiques
| Révision :

root / man / man3 / HPL_pdrpanllT.3

Historique | Voir | Annoter | Télécharger (2,58 ko)

1 
.TH HPL_pdrpanllT 3 "September 10, 2008" "HPL 2.0" "HPL Library Functions"
2 
.SH NAME
3 
HPL_pdrpanllT \- Left-looking recursive panel factorization.
4 
.SH SYNOPSIS
5 
\fB\&#include "hpl.h"\fR
6
7 
\fB\&void\fR
8 
\fB\&HPL_pdrpanllT(\fR
9 
\fB\&HPL_T_panel *\fR
10 
\fI\&PANEL\fR,
11 
\fB\&const int\fR
12 
\fI\&M\fR,
13 
\fB\&const int\fR
14 
\fI\&N\fR,
15 
\fB\&const int\fR
16 
\fI\&ICOFF\fR,
17 
\fB\&double *\fR
18 
\fI\&WORK\fR
19 
\fB\&);\fR
20 
.SH DESCRIPTION
21 
\fB\&HPL_pdrpanllT\fR
22 
recursively  factorizes  a panel of columns  using  the
23 
recursive Left-looking variant of the one-dimensional algorithm.  The
24 
lower  triangular  N0-by-N0  upper block  of  the panel  is stored in
25 
transpose form.
26
27 
Bi-directional  exchange  is  used  to  perform  the  swap::broadcast
28 
operations  at once  for one column in the panel.  This  results in a
29 
lower number of slightly larger  messages than usual.  On P processes
30 
and assuming bi-directional links,  the running time of this function
31 
can be approximated by (when N is equal to N0):
32
33 
   N0 * log_2( P ) * ( lat + ( 2*N0 + 4 ) / bdwth ) +
34 
   N0^2 * ( M - N0/3 ) * gam2-3
35
36 
where M is the local number of rows of  the panel, lat and bdwth  are
37 
the latency and bandwidth of the network for  double  precision  real
38 
words,  and  gam2-3  is an estimate of the  Level 2 and Level 3  BLAS
39 
rate of execution. The  recursive  algorithm  allows indeed to almost
40 
achieve  Level 3 BLAS  performance  in the panel factorization.  On a
41 
large  number of modern machines,  this  operation is however latency
42 
bound,  meaning  that its cost can  be estimated  by only the latency
43 
portion N0 * log_2(P) * lat.  Mono-directional links will double this
44 
communication cost.
45 
.SH ARGUMENTS
46 
.TP 8
47 
PANEL   (local input/output)    HPL_T_panel *
48 
On entry,  PANEL  points to the data structure containing the
49 
panel information.
50 
.TP 8
51 
M       (local input)           const int
52 
On entry,  M specifies the local number of rows of sub(A).
53 
.TP 8
54 
N       (local input)           const int
55 
On entry,  N specifies the local number of columns of sub(A).
56 
.TP 8
57 
ICOFF   (global input)          const int
58 
On entry, ICOFF specifies the row and column offset of sub(A)
59 
in A.
60 
.TP 8
61 
WORK    (local workspace)       double *
62 
On entry, WORK  is a workarray of size at least 2*(4+2*N0).
63 
.SH SEE ALSO
64 
.BR HPL_dlocmax \ (3),
65 
.BR HPL_dlocswpN \ (3),
66 
.BR HPL_dlocswpT \ (3),
67 
.BR HPL_pdmxswp \ (3),
68 
.BR HPL_pdpancrN \ (3),
69 
.BR HPL_pdpancrT \ (3),
70 
.BR HPL_pdpanllN \ (3),
71 
.BR HPL_pdpanllT \ (3),
72 
.BR HPL_pdpanrlN \ (3),
73 
.BR HPL_pdpanrlT \ (3),
74 
.BR HPL_pdrpancrN \ (3),
75 
.BR HPL_pdrpancrT \ (3),
76 
.BR HPL_pdrpanllN \ (3),
77 
.BR HPL_pdrpanrlN \ (3),
78 
.BR HPL_pdrpanrlT \ (3),
79 
.BR HPL_pdfact \ (3).