more timing
"Eric S. Johansson" <[email protected]>
| Newsgroups | gmane.mail.spam.crm114 |
|---|---|
| Message-ID | <[email protected]> |
ran the same test on a real machine and got interesting results. the 4x parallel version ran 1.75 times as long as the 1x pass. implication: on a real machine running 2 crm114 processes is a win but on a vm, a single crm114 process is the highest throughput. still am wondering why my scores are so far off but that is a different discussion. timings as before 0.0, 11065 1.1, 1 0.2, 2 0.3, 1 0.6, 1 0.1, 36 0.7, 1 real 19m19.722s user 8m58.026s sys 2m24.161s ---------- 0.0, 9962 1.0, 1 0.1, 945 0.2, 163 0.3, 31 0.6, 2 0.4, 3 real 34m25.418s user 8m47.085s sys 2m15.788s 0.0, 9931 0.2, 174 0.4, 2 0.3, 26 0.6, 1 0.1, 973 real 34m17.557s user 8m45.817s sys 2m12.664s 0.0, 9955 0.5, 1 0.2, 160 0.4, 1 0.3, 20 0.1, 970 real 34m20.348s user 8m47.573s sys 2m13.204s 0.0, 9958 0.5, 1 0.2, 149 0.4, 1 0.3, 24 0.1, 974 real 34m23.236s user 8m44.433s sys 2m12.868s ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php