|
|
Glas :Re: [glas] performance tests |
From: Karl Meerbergen (Karl.Meerbergen_at_[hidden])
Date: 2005-11-02 03:41:43
Good job!
Just one remark for the complex_double benchmark. ublas does not implement the
norm_1, neither does blas's DZASUM. So, some caution is required here with
the timings. Timings should be done for glas::abs_sum() rather than
glas::norm_1().
Karl
On Monday 31 October 2005 21:39, Toon Knapen wrote:
> Hi all,
>
> Because performance is of utmost importance for GLAS algorithms (after
> correctness of course), I have been working on some basic infrastructure
> to test the performance of the different algorithms and visualise the
> results. These graphs will show the time of the different operations
> _ for different sizes of the input arguments
> _ compared to other libraries
> _ on different types of machines.
>
> I just updated the website and so far results are available for the
> norm_1 algorithm on
>
> vectors of double:
> http://glas.sourceforge.net/perf_bench/norm_1/double/index.html
>
> vectors of complex:
> http://glas.sourceforge.net/perf_bench/norm_1/complex_double/index.html
>
>
> On the home-page there is also a link to 'Performance tests'. At the
> bottom of that page there is a link to the overview of performance tests
> that have executed.
>
> Any feedback is very welcome,
>
> toon
-- ============================================== Look at our unique training program and Register on-line at http://www.fft.be/?id=35 ---------------------------------------------- Karl Meerbergen Free Field Technologies 16 place de l'Université B-1348 Louvain-la-Neuve - BELGIUM Company Phone: +32 10 45 12 26 Company Fax: +32 10 45 46 26 Mobile Phone: +32 474 26 66 59 Home Phone: +32 2 306 38 10 mailto:Karl.Meerbergen_at_[hidden] http://www.fft.be ==============================================