Hi, Am Samstag 12 Februar 2005 20.47 schrieb Steven M. Schultz: > The final total elapsed time of the two encoding runs: > > dual G5: 34hrs 7min 1.3 frames/sec > dual Opteron: 15hrs 46 min 2.80 frames/sec > I count y4mdenoise an excellent piece of software but I thought that I make something wrong: On my machine (1.2GHz Athlon) I obtain a denoising rate of 0.3 frames/sec. (that is, 5 days for a movie) This is far too slow for routine purposes. Or, the other way round, even a small increase in speed would give a lot. Therefore my question: Does anyone has experiences with optimization? The optimization switches in the CVS build are rather moderate (-O2 -mach=athlon-tbird). Are there any experiences using other compilers like icc? Another point: Having a look at the implementation document (I must admit that I did not understand much) indicates that there are rather complex data structures involved. Are there any attempts to adapt them to the memory hierarchy of modern processor architectures? (see for example the ATLAS project for the BLAS) A well-tuned data access pattern can boost the performance a lot.
This is something like a dream... Michael ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Mjpeg-users mailing list Mjpeg-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mjpeg-users
