| 1 |
<HTML> |
| 2 |
<HEAD> |
| 3 |
<TITLE>HPL_pdrpanllN HPL 2.0 Library Functions September 10, 2008</TITLE> |
| 4 |
</HEAD> |
| 5 |
|
| 6 |
<BODY BGCOLOR="WHITE" TEXT = "#000000" LINK = "#0000ff" VLINK = "#000099" |
| 7 |
ALINK = "#ffff00"> |
| 8 |
|
| 9 |
<H1>Name</H1> |
| 10 |
<B>HPL_pdrpanllN</B> Left-looking recursive panel factorization. |
| 11 |
|
| 12 |
<H1>Synopsis</H1> |
| 13 |
<CODE>#include "hpl.h"</CODE><BR><BR> |
| 14 |
<CODE>void</CODE> |
| 15 |
<CODE>HPL_pdrpanllN(</CODE> |
| 16 |
<CODE>HPL_T_panel *</CODE> |
| 17 |
<CODE>PANEL</CODE>, |
| 18 |
<CODE>const int</CODE> |
| 19 |
<CODE>M</CODE>, |
| 20 |
<CODE>const int</CODE> |
| 21 |
<CODE>N</CODE>, |
| 22 |
<CODE>const int</CODE> |
| 23 |
<CODE>ICOFF</CODE>, |
| 24 |
<CODE>double *</CODE> |
| 25 |
<CODE>WORK</CODE> |
| 26 |
<CODE>);</CODE> |
| 27 |
|
| 28 |
<H1>Description</H1> |
| 29 |
<B>HPL_pdrpanllN</B> |
| 30 |
recursively factorizes a panel of columns using the |
| 31 |
recursive Left-looking variant of the one-dimensional algorithm. The |
| 32 |
lower triangular N0-by-N0 upper block of the panel is stored in |
| 33 |
no-transpose form (i.e. just like the input matrix itself). |
| 34 |
|
| 35 |
Bi-directional exchange is used to perform the swap::broadcast |
| 36 |
operations at once for one column in the panel. This results in a |
| 37 |
lower number of slightly larger messages than usual. On P processes |
| 38 |
and assuming bi-directional links, the running time of this function |
| 39 |
can be approximated by (when N is equal to N0): |
| 40 |
|
| 41 |
N0 * log_2( P ) * ( lat + ( 2*N0 + 4 ) / bdwth ) + |
| 42 |
N0^2 * ( M - N0/3 ) * gam2-3 |
| 43 |
|
| 44 |
where M is the local number of rows of the panel, lat and bdwth are |
| 45 |
the latency and bandwidth of the network for double precision real |
| 46 |
words, and gam2-3 is an estimate of the Level 2 and Level 3 BLAS |
| 47 |
rate of execution. The recursive algorithm allows indeed to almost |
| 48 |
achieve Level 3 BLAS performance in the panel factorization. On a |
| 49 |
large number of modern machines, this operation is however latency |
| 50 |
bound, meaning that its cost can be estimated by only the latency |
| 51 |
portion N0 * log_2(P) * lat. Mono-directional links will double this |
| 52 |
communication cost. |
| 53 |
|
| 54 |
<H1>Arguments</H1> |
| 55 |
<PRE> |
| 56 |
PANEL (local input/output) HPL_T_panel * |
| 57 |
On entry, PANEL points to the data structure containing the |
| 58 |
panel information. |
| 59 |
</PRE> |
| 60 |
<PRE> |
| 61 |
M (local input) const int |
| 62 |
On entry, M specifies the local number of rows of sub(A). |
| 63 |
</PRE> |
| 64 |
<PRE> |
| 65 |
N (local input) const int |
| 66 |
On entry, N specifies the local number of columns of sub(A). |
| 67 |
</PRE> |
| 68 |
<PRE> |
| 69 |
ICOFF (global input) const int |
| 70 |
On entry, ICOFF specifies the row and column offset of sub(A) |
| 71 |
in A. |
| 72 |
</PRE> |
| 73 |
<PRE> |
| 74 |
WORK (local workspace) double * |
| 75 |
On entry, WORK is a workarray of size at least 2*(4+2*N0). |
| 76 |
</PRE> |
| 77 |
|
| 78 |
<H1>See Also</H1> |
| 79 |
<A HREF="HPL_dlocmax.html">HPL_dlocmax</A>, |
| 80 |
<A HREF="HPL_dlocswpN.html">HPL_dlocswpN</A>, |
| 81 |
<A HREF="HPL_dlocswpT.html">HPL_dlocswpT</A>, |
| 82 |
<A HREF="HPL_pdmxswp.html">HPL_pdmxswp</A>, |
| 83 |
<A HREF="HPL_pdpancrN.html">HPL_pdpancrN</A>, |
| 84 |
<A HREF="HPL_pdpancrT.html">HPL_pdpancrT</A>, |
| 85 |
<A HREF="HPL_pdpanllN.html">HPL_pdpanllN</A>, |
| 86 |
<A HREF="HPL_pdpanllT.html">HPL_pdpanllT</A>, |
| 87 |
<A HREF="HPL_pdpanrlN.html">HPL_pdpanrlN</A>, |
| 88 |
<A HREF="HPL_pdpanrlT.html">HPL_pdpanrlT</A>, |
| 89 |
<A HREF="HPL_pdrpancrN.html">HPL_pdrpancrN</A>, |
| 90 |
<A HREF="HPL_pdrpancrT.html">HPL_pdrpancrT</A>, |
| 91 |
<A HREF="HPL_pdrpanllT.html">HPL_pdrpanllT</A>, |
| 92 |
<A HREF="HPL_pdrpanrlN.html">HPL_pdrpanrlN</A>, |
| 93 |
<A HREF="HPL_pdrpanrlT.html">HPL_pdrpanrlT</A>, |
| 94 |
<A HREF="HPL_pdfact.html">HPL_pdfact</A>. |
| 95 |
|
| 96 |
</BODY> |
| 97 |
</HTML> |