On 08/17/2012 06:29 PM, Roy Stogner wrote: > > On Fri, 17 Aug 2012, Jens Lohne Eftang wrote: > >> That helps a lot. Now I also do the static allocation, but in >> addition reducing io_blksize to 2560 knocked another 180 seconds off >> the computation time. >> >> The time for reading in the 1GB of data is now 20 seconds (all calls >> to read_serialized_vector), which I still find a bit too much, but >> obviously way better than before. Thanks! > > Something's not adding up (subtracting up?) for me - 360 seconds > originally, minus 80 seconds with static allocation, minus 180 seconds > with reduced block size, gives me 100 seconds left, not 20. > The 360 was the whole program execution time. The 20 seconds is the data read only. Sorry about the confusion. > Also: what does the time come out to if you do the reduced block size > but not the static allocation? That's the case that I'm particularly > interested in, since we could commit something approximating it to > svn. That did not change much. Without the static allocation reading the data is 20.5 secs.
------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
