Re: GMT output Format
"Eduardo A. Suárez" <[email protected]>
| Newsgroups | gmane.comp.gis.gmt.user |
|---|---|
| Message-ID | <[email protected]> |
My setup is this Core2 Duo w/ 8 GB RAM and HDD (no SSD). Debian GNU/Linux testing (jessie), 32 bits. I'm using a 2G tmpfs [1] ramdisk and forcing a cache flush [2] prior to each test. time grdinfo etopo1m_grd.nc always gives between 6-9 seg (this a busy machine, so it depends on machine load), even when I mistakenly issued the command in a wrong directory (file not found error). Without flushing the cache the time is 0.07 seg I think all the time is spent loading the gmt executable and needed libraries. Eduardo.- [1] http://kvz.io/blog/2007/07/18/create-turbocharged-storage-using-tmpfs/ [2] http://catalin-festila.blogspot.com.ar/2011/06/myth-of-dropcaches.html Quoting Remko Scharroo <[email protected]>: > But the copying of the file will have put it in cache, so that would > make it go faster. > You would have to do the copy first, then come back much later to do > the grdinfo. > > Remko > > On 21 Jan 2014, at 18:33, Paul Wessel <[email protected]> wrote: > >> Good point, of course. My quick test of making a RAM disk under >> OSX, place a copy of etopo1m.nc there (890 Mb) and run time grdinfo >> etopo1m.nc was disappointing; it took about the same 8 seconds to >> finish as it does on the HD.... >> >> -p > > To unsubscribe, send the message "signoff gmt-help" to > [email protected] > -- Eduardo A. Suarez Facultad de Ciencias Astronómicas y Geofísicas - UNLP FCAG: (0221)-4236593 int. 172/Cel: (0221)-15-4557542/Casa: (0221)-4526589 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. To unsubscribe, send the message "signoff gmt-help" to [email protected]