[Beowulf] Benchmarking?
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Rajiv jrajiv at hclinsys.comMon Aug 2 20:27:41 PDT 2004
- Previous message: [Beowulf] EM64T Clusters
- Next message: [Beowulf] Running Matlab and Femlab in Batch Mode on a beowulf
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Dear Michael,
I am sending below the output. Also find the attachment HPL.dat.
OUTPUT-1:
If my machines file has one entry client, the following is the output:
[hpl at master hpl]$ /opt/mpich/gnu/bin/mpirun -nolocal -np 2 -machinefile
machines /opt/hpl/gnu/bin/xhpl Warning: No xauth data; using fake
authentication data for X11 forwarding.
============================================================================
HPLinpack 1.0 -- High-Performance Linpack benchmark -- September 27,
2000
Written by A. Petitet and R. Clint Whaley, Innovative Computing Labs., UTK
============================================================================
An explanation of the input/output parameters follows:
T/V : Wall time / encoded variant.
N : The order of the coefficient matrix A.
NB : The partitioning blocking factor.
P : The number of process rows.
Q : The number of process columns.
Time : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.
The following parameter values will be used:
N : 1000
NB : 64
P : 1
Q : 2
PFACT : Left Crout Right
NBMIN : 8
NDIV : 2
RFACT : Right
BCAST : 1ringM
DEPTH : 1
SWAP : Mix (threshold = 80)
L1 : transposed form
U : transposed form
EQUIL : yes
ALIGN : 8 double precision words
----------------------------------------------------------------------------
- The matrix A is randomly generated for each test.
- The following scaled residual checks will be computed:
1) ||Ax-b||_oo / ( eps * ||A||_1 * N )
2) ||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 )
3) ||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo )
- The relative machine precision (eps) is taken to be 1.110223e-16
- Computational tests pass if scaled residuals are less than 16.0
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2L8 1000 64 1 2 0.29 2.294e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 1.0108964 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0245358 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0059277 ...... PASSED
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2C8 1000 64 1 2 0.26 2.617e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 1.0164336 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0246702 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0059601 ...... PASSED
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2R8 1000 64 1 2 0.26 2.592e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 0.9860995 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0239340 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0057823 ...... PASSED
============================================================================
Finished 3 tests with the following results:
3 tests completed and passed residual checks,
0 tests completed and failed residual checks,
0 tests skipped because of illegal input values.
----------------------------------------------------------------------------
End of Tests.
============================================================================
OUTPUT-2:
Output with two entries in the machine file, master and client
[hpl at master hpl]$ /opt/mpich/gnu/bin/mpirun -nolocal -np 4 -machinefile
machines /opt/hpl/gnu/bin/xhpl
Warning: No xauth data; using fake authentication data for X11 forwarding.
============================================================================
HPLinpack 1.0 -- High-Performance Linpack benchmark -- September 27,
2000
Written by A. Petitet and R. Clint Whaley, Innovative Computing Labs., UTK
============================================================================
An explanation of the input/output parameters follows:
T/V : Wall time / encoded variant.
N : The order of the coefficient matrix A.
NB : The partitioning blocking factor.
P : The number of process rows.
Q : The number of process columns.
Time : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.
The following parameter values will be used:
N : 1000
NB : 64
P : 1
Q : 2
PFACT : Left Crout Right
NBMIN : 8
NDIV : 2
RFACT : Right
BCAST : 1ringM
DEPTH : 1
SWAP : Mix (threshold = 80)
L1 : transposed form
U : transposed form
EQUIL : yes
ALIGN : 8 double precision words
----------------------------------------------------------------------------
- The matrix A is randomly generated for each test.
- The following scaled residual checks will be computed:
1) ||Ax-b||_oo / ( eps * ||A||_1 * N )
2) ||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 )
3) ||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo )
- The relative machine precision (eps) is taken to be 1.110223e-16
- Computational tests pass if scaled residuals are less than 16.0
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2L8 1000 64 1 2 0.38 1.775e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 1.0108964 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0245358 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0059277 ...... PASSED
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2C8 1000 64 1 2 0.37 1.807e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 1.0525457 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0255467 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0061719 ...... PASSED
============================================================================
T/V N NB P Q Time Gflops
----------------------------------------------------------------------------
W11R2R8 1000 64 1 2 0.39 1.692e+00
----------------------------------------------------------------------------
||Ax-b||_oo / ( eps * ||A||_1 * N ) = 1.0908244 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_1 * ||x||_1 ) = 0.0264758 ...... PASSED
||Ax-b||_oo / ( eps * ||A||_oo * ||x||_oo ) = 0.0063964 ...... PASSED
============================================================================
Finished 3 tests with the following results:
3 tests completed and passed residual checks,
0 tests completed and failed residual checks,
0 tests skipped because of illegal input values.
For this outputs could you please tell me what is the performance. Also tell
me what is this message "Warning: No xauth data; using fake authentication
data for X11 forwarding".
Regards,
Rajiv
----- Original Message -----
From: "Rajiv" <jrajiv at hclinsys.com>
To: <mwill at penguincomputing.com>; <beowulf at beowulf.org>
Sent: Thursday, July 29, 2004 1:53 PM
Subject: Re: [Beowulf] Benchmarking?
> Dear Michael,
> I have installed HPL and is working fine. After executing
> $ /opt/mpich/gnu/bin/mpirun -nolocal -np 4 -machinefile machines
> /opt/hpl/gnu/bin/xhpl
>
> I get three results. What are those results. Can I say the maximum among
> these three result (in GFLOPS) is the performance of the cluster.
>
> Regards,
> Rajiv
>
> ----- Original Message -----
> From: "Michael Will" <mwill at penguincomputing.com>
> To: <beowulf at beowulf.org>
> Sent: Thursday, July 29, 2004 2:13 AM
> Subject: Re: [Beowulf] Benchmarking?
>
>
> > description and link for download on
http://www.netlib.org/benchmark/hpl/
> >
> > Michael
> > On Wednesday 28 July 2004 11:57 am, Jack C wrote:
> > > I was just told about linpack, and that it was good. And apperantly
> there is
> > > an MPI implementation thereof, but I sure can't find it... the MPI
> standard
> > > site is not very informative. :(
> > >
> > > Anyone know where to find that, or any other good ones?
> > >
> > > Thanks again,
> > >
> > > -Jack C
> > > http://www.CREPinc.com
> > > _______________________________________________
> > > Beowulf mailing list, Beowulf at beowulf.org
> > > To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf
> > >
> >
> > --
> > Michael Will, Linux Sales Engineer
> > NEWS: We have moved to a larger iceberg :-)
> > NEWS: 300 California St., San Francisco, CA.
> > Tel: 415-954-2822 Toll Free: 888-PENGUIN
> > Fax: 415-954-2899
> > www.penguincomputing.com
> >
> > _______________________________________________
> > Beowulf mailing list, Beowulf at beowulf.org
> > To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit
http://www.beowulf.org/mailman/listinfo/beowulf
-------------- next part --------------
A non-text attachment was scrubbed...
Name: HPL.dat
Type: application/octet-stream
Size: 1039 bytes
Desc: not available
Url : http://www.scyld.com/pipermail/beowulf/attachments/20040803/eab79c7a/HPL.obj
- Previous message: [Beowulf] EM64T Clusters
- Next message: [Beowulf] Running Matlab and Femlab in Batch Mode on a beowulf
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list