发信人: zxs()
整理人: zhcharles(2002-01-30 15:49:00), 站内信件
|
CPU: intel's PII233
1) MS PWinNt 4.0, Vc 5.0,
-O2,
FLOPS C Program (Double Precision), V2.0 18 Dec 1992
Module Error RunTime MFLOPS
(usec)
1 2.8422e-014 0.1596 87.7141
2 2.5047e-013 0.1123 62.3521
3 -7.6605e-015 0.1692 100.4524
4 2.2771e-013 0.1669 89.8876
5 3.8858e-014 0.3281 88.3788
6 7.5495e-015 0.2441 118.8030
7 -1.1369e-013 0.4143 28.9669
8 1.2612e-013 0.2747 109.2150
Iterations = 128000000
NullTime (usec) = 0.0000
MFLOPS(1) = 71.1780
MFLOPS(2) = 55.7795
MFLOPS(3) = 83.1006
MFLOPS(4) = 106.4454
2)
Linux 2.0.30, gcc 2.7
-O -O2
FLOPS C Program (Double Precision), V2.0 18 Dec 1992
Module Error RunTime MFLOPS
(usec)
1 -5.4193e-13 0.1592 87.9293
2 8.5760e-16 0.1322 52.9551
3 3.4567e-14 0.1612 105.4775
4 3.6970e-13 0.1722 87.1143
5 -5.1910e-15 0.4029 71.9798
6 2.3930e-14 0.2845 101.9500
7 -1.6524e-10 0.4770 25.1597
8 1.4631e-13 0.2941 101.9920
Iterations = 128000000
NullTime (usec) = 0.0087
MFLOPS(1) = 63.2519
MFLOPS(2) = 49.2270
MFLOPS(3) = 74.8328
MFLOPS(4) = 99.7858
3) Linux 2.0.30, gcc 2.7
-O -O6 (although the manual only refers -O3 as the
fastest optimization, I still try the O6 option
Mr. Zhengzy's recommendation)
FLOPS C Program (Double Precision), V2.0 18 Dec 1992
Module Error RunTime MFLOPS
(usec)
1 -5.4193e-13 0.1594 87.8431
2 8.5760e-16 0.1324 52.8614
3 3.4567e-14 0.1608 105.7337
4 3.6970e-13 0.1720 87.1935
5 -5.1910e-15 0.4031 71.9380
6 2.3930e-14 0.2840 102.1183
7 -1.6524e-10 0.4772 25.1473
8 1.4631e-13 0.2943 101.9379
Iterations = 128000000
NullTime (usec) = 0.0087
MFLOPS(1) = 63.1919
MFLOPS(2) = 49.2221
MFLOPS(3) = 74.8418
MFLOPS(4) = 99.8800
As a result, my PII outperforms many old workstations,
such as HP9000/735, SGI's R8000, IBM's R6000,
But..., apparently its capacity does not be utilized
fully.
--
m7m※ 来源:·BBS 水木清华站 bbs.net.tsinghua.edu.cn·[FROM: 202.120.26.50]m
|
|