*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
I have made some initial experiments to exploit the dual core CPUs for crystallographic calculations via OpenMP with rather disappointing results. What is however very useful and simple to do (by hand or with a script) is to run SHELXD several times in parallel to speed up the heavy atom location. If there is no SEED instruction in the .ins file, the program will use a different random number seed each time even when the files are identical, though of course one has to use different filenames. On a machine with two dual core Xeons, ie. 4 CPUs, four SHELXD jobs run in exactly the same real time as one. It is even better to run 8 at once to take advantage of the hyperthreading, this gives the best overall throughput. Running 5 jobs at once is less effective because (unlike the 4 and 8 cases) they do not finish at the same time! It would be very easy to implement this in the various GUIs that call SHELXD, but as far as I know it has not been done yet. An alternative strategy for difficult cases would be to run 4 or 8 jobs simultaneously with different resolution cutoffs to see which is best. There is one proviso, the computer must have enough memory (not usually a problem with SHELXD).
George -- Prof. George M. Sheldrick FRS Dept. Structural Chemistry, University of Goettingen, Tammannstr. 4, D37077 Goettingen, Germany Tel. +49-551-39-3021 or -3068 Fax. +49-551-39-2582
