Next Step...

到目前為止，SMP還是跑在Linux上比較好，理由是分數比較高（這算理由嗎？），相關測試數據可參考這篇文章：Folding on the Intel V8 - Eight Cores of F@H

Justin West用8核心硬體做了四組測試：
1.在Windows XP 32bit上跑兩個SMP Client
2.在Windows XP 32bit上用VMWare跑4個Ubuntu 64bit，而後在這4個Ubuntu上跑1個Linux SMP Client
3.用三個VMWare上的Ubuntu 64bit跑Linux SMP Client，並多跑一個GPU Client，該顯卡為x1950XTX
4.在Ubuntu 64bit上跑兩個Linux SMP Client

而分數分別為：
1=3980
2=4530
3=3250
4=4668

個人不太會評估這測試的公正性，不過該作者最後說：
if you want to get the most out of your system, run Linux!
所以還是用Linux跑好了～XD

又或者～參考這裡的介紹：Folding@Home SMP Client with VMware & Ubuntu

在Windows上裝個VMWare跑ubuntu～

在Linux上的安裝方法相當簡單，連到：

http://folding.stanford.edu/English/Download

下載Linux的版本

建構一個資料夾，將下載的檔案解壓縮進去。
解壓縮後有兩個檔案：fah6與mpiexec
直接執行： ./fah6 -smp
照著與windows相同的設定，就會開始執行了

比較特別的地方是，Linux在執行之後會有些運算結果數據～列表如下：

M E G A - F L O P S A C C O U N T I N G Parallel run - timing based on wallclock. RF=Reaction-Field FE=Free Energy SCFE=Soft-Core/Free Energy T=Tabulated W3=SPC/TIP3p W4=TIP4p (single or pairs) NF=No Forces Computing: M-Number M-Flops % of Flops ----------------------------------------------------------------------- VdW(T) 40039.749431 2162146.469274 0.9 RF Coul 254373.868216 8394337.651128 3.6 RF Coul [W3] 3310.364494 324415.720412 0.1 RF Coul + VdW(T) 717152.229091 46614894.890915 20.2 RF Coul + VdW(T) [W3] 115948.638570 15073323.014100 6.5 RF Coul + VdW(T) [W3-W3] 448631.074742 143561943.917440 62.2 Outer nonbonded loop 58508.577060 585085.770600 0.3 1,4 nonbonded interactions 7279.102428 655119.218520 0.3 NS-Pairs 344466.979877 7233806.577417 3.1 Reset In Box 1569.694416 14127.249744 0.0 Shift-X 31336.509048 188019.054288 0.1 CG-CoM 707.764848 20525.180592 0.0 Sum Forces 47085.477192 47085.477192 0.0 Bonds 1431.638052 61560.436236 0.0 Angles 5015.006256 817446.019728 0.4 RB-Dihedrals 6363.398496 1571759.428512 0.7 Virial 15706.553928 282717.970704 0.1 Update 15695.159064 486549.930984 0.2 Stop-CM 15695.010306 156950.103060 0.1 P-Coupling 15695.159064 94170.954384 0.0 Calc-Ekin 15695.307822 423773.311194 0.2 Lincs 1345.543524 80732.611440 0.0 Lincs-Mat 6510.265632 26041.062528 0.0 Constraint-V 15695.159064 94170.954384 0.0 Constraint-Vir 14273.016732 342552.401568 0.1 Settle 4309.157736 1391857.948728 0.6 ----------------------------------------------------------------------- Total 230705113.325072 100.0 ----------------------------------------------------------------------- NODE (s) Real (s) (%) Time: 47753.000 47753.000 100.0 13h15:53 (Mnbf/s) (GFlops) (ns/day) (hour/ns) Performance: 33.076 4.831 0.905 26.529