Fwd: Re: Slow Sfepy on Fedoracore 8
Hi,
I did not send the reply to the list - correcting now.
Also try running
$ ./runTests.py tests/test_linear_solvers.py --filter-less
to see the performance of available solvers applied to a Laplace problem.
For the reference, on my slow old laptop I get:
... solution times (rezidual norms): ... 1.41 [s] (3.286e-12) : i20 ls.scipy_iterative cg ... 2.25 [s] (1.714e-12) : i21 ls.scipy_iterative bicgstab ... 5.37 [s] (3.575e-12) : i22 ls.scipy_iterative qmr ... 12.92 [s] (4.135e-15) : d00 ls.umfpack ... 10000000000.00 [s] (1.000e+10) : i00 ls.pyamg ruge_stuben_solver ... 10000000000.00 [s] (1.000e+10) : i01 ls.pyamg smoothed_aggregation_solver ... 10000000000.00 [s] (1.000e+10) : i10 ls.petsc cg icc
The last three lines mean that pyamg and petsc are not installed. BTW. I recommend you to install petsc and petsc4py, as the iterative solvers and preconditioners in petsc are really good.
r.
----- Forwarded message from cimr...@ntc.zcu.cz ----- Date: Sat, 25 Jul 2009 11:42:23 +0200 From: Robert Cimrman cimr...@ntc.zcu.cz Reply-To: Robert Cimrman cimr...@ntc.zcu.cz Subject: Re: Slow Sfepy on Fedoracore 8 To: Antonino Ingargiola trit...@gmail.com
Quoting Antonino Ingargiola trit...@gmail.com:
2009/7/23 Robert Cimrman cimr...@ntc.zcu.cz: <snip> On my machine (a quad-core 64bit, gentoo), all tests
pass in:
32 test file(s) executed in 41.48 s, 0 failure(s) of 42 test(s)
you should get similar results. In fact on a dual core ubuntu machine I run all the test in less
then a minute.On fedoracore8 I don't know how slow is but surely more than 20min (I interrupted the tests) that is definitely too much. It definitely looks like umfpack is not used.
Is possible that is another library that is not used causing the long runtime? For example atlas/lapack?
Maybe. I am really not sure what happens.
How can I check if scipy is using atlas/lapack/umfpack? Maybe I should ask on the scipy mailing list....
Yes, I have seen the thread - so it seems ATLAS (blas+lapack) is ok.
But it looks like the umfpack itself might not use it - can you try to compile it manually? Otherwise I cannot explain its poor performance.
r.
----- End forwarded message -----
participants (1)
-
Robert Cimrman