Is there any advantage to using say dual Xeon 2.8GHz nodes with DDR266 over twice as many nodes with P4s and DDR400 memory?
We run simulation software that uses masses of indirect addressing and requires frequent summation over all node domains and have tentatively opted for the single cpu nodes. Can anyone point me to some literature discussing these or similar issues? Also, I am looking for tools to optimise codes for cluster performance. Any pointers welcome. Eugene ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click _______________________________________________ Oscar-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/oscar-users
