site_archiver@lists.apple.com Delivered-To: darwin-dev@lists.apple.com On Sep 25, 2009, at 1:29 PM, Tomasz Koziara wrote: For each time t For each output file f Seek (f, t) Read (f) —Jens _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... The post-processing process (single-thread) simply opens all 3 * N files and keeps them open. Of those 3 * N files, only N store some substantial data (are large), while the rest stores indexing and labeling data. So basically what happens is: Sounds like you're reading a lot of data, so I would expect that your process spends most of its time blocked in disk I/O, which is what the profiles show. It sounds like your access patterns are pretty random too, which makes performance worse. I don't know anything at all about MPI or XDR, so I don't think I can give any advice here. This email sent to site_archiver@lists.apple.com