
Are you sure there aren't any 32-bit shared libraries loaded? Because if there are, then you will save a lot of memory by using the 32-bit version of chimera. Also the 32-bit version of chimera is faster because it pushs around less memory (the data structures are smaller). The benchmark scores are inline with what one would expect from an integrated graphics solution. If you want higher scores, you need to use ATI or NVidia graphics. And yes, chimera is not multi-threaded. There are some computations within chimera that would benefit, but the most benefit comes from when OpenGL is multi-threaded and the newer graphics drivers are (NVidia, Apple today, more coming). Greg Couch UCSF Computer Graphics Lab On Tue, 21 Aug 2007, Damir Perisa wrote:
hi there,
i just set up my new laptop with archlinux 64bit and since i'm going to use everything 64bit on this machine (no 32bit libraries), i tried first time chimera 64bit.
details on machine: Thinkpad x61t (Tablet Laptop) Linux Apollon 2.6.22-ARCH #1 SMP PREEMPT Wed Aug 15 22:51:57 CEST 2007 x86_64 Intel(R) Core(TM)2 Duo CPU L7500 @ 1.60GHz GenuineIntel GNU/Linux
graphics: Intel integrated graphics 965GM VideoRam: 262144 KB (xorg version 1.3.0 xf86-video-intel version 2.1.1)
ram total: [damir@Apollon ~]$ free total used free shared buffers cached Mem: 1002364 570780 431584 0 9524 142080 -/+ buffers/cache: 419176 583188 Swap: 979924 372324 607600
benchmarks: surface 130 mesh 129 contour 158 solid 128 recolor 129
Molecule (Atoms): Null Wire Stick BStick Ribbon Sphere Ops 1GCN (246): 762.0 677.0 150.0 150.0 736.0 109.0 7634.0 1YTI (798): 981.0 554.0 50.0 48.0 702.0 36.0 2305.0 2GBP (2575): 877.0 375.0 17.0 17.0 609.0 16.0 713.0 1GOT (6449): 601.0 213.0 6.0 6.0 468.0 9.0 216.0 2HMG (12189): 254.0 134.0 3.0 3.0 291.0 6.0 87.0 2PDA (18894): 165.0 91.0 2.0 2.0 191.0 4.0 51.0 1YNJ (24369): 132.0 67.0 5.0 10.0 217.0 3.0 40.0 1F4H (33805): 97.0 53.0 1.0 1.0 128.0 2.0 28.0
quite low scores, no? ... in addition i'm observing that only 50% of cpu is used (means only one core). chimera is not multi-threaded?
greetings, Damir _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users