Benchmark Results for <I>pheat.module</I>

Benchmark Results for pheat.module

(For questions about these benchmarks send eMail to Claudio Fleiner)

This program (pheat.sa) calculates in a very naive manner how a flat surface heats up when the four corners are heated. For this benchmark we used a square matrix of size 2000 and calculated one step. The heat of a new cell is the mean of the heat of itself and all its 8 neighbors. The program ran on a 4 processor Sparc in parallel.

Comparsion of pSather and Sather

Programoverall exection timeinitializationcalculation onlyspeedup
Sather25.75.620.1
pSather12.27.15.13.94

Table of Contents

General Results:

# [#] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE *sec. *% of
std
KB Shorter
in %
sec. % of
std
KB % of
std
1. Standard Compiler with no options 44.70100.0%3200.0%13.52100.0%7798.10100.0%
2. [30] 13.3029.8%17644.9%22.55166.8%10699.43137.2%
3. [18] 13.3029.8%17644.9%22.46166.1%11780.92151.1%
4. [26] 13.3029.8%17744.7%23.20171.6%12615.68161.8%
5. [22] 13.3029.8%17545.1%21.80161.2%10437.97133.9%
6. [27] 13.3029.8%17744.7%23.17171.4%11802.62151.4%
7. [19] 13.3029.8%17644.9%22.18164.1%11931.65153.0%
8. [23] 13.4030.0%17545.1%21.88161.8%11916.63152.8%
9. [31]13.4030.0%17644.9%22.60167.2%11915.26152.8%
10. [29] 13.9031.1%17644.9%19.62145.1%9800.09125.7%
11. [20] 13.9031.1%17545.1%19.06141.0%10434.15133.8%
12. [21] 13.9031.1%17545.1%19.18141.9%9723.08124.7%
13. [28] 13.9031.1%17644.9%19.75146.1%10398.92133.4%
14. [25] 13.9031.1%17644.8%19.55144.6%10115.48129.7%
15. [16] 14.0031.3%17645.0%18.99140.5%10448.08134.0%
16. [17] 14.0031.3%17645.0%18.79139.0%9681.58124.2%
17. [24] 14.1031.5%17644.8%19.69145.6%10486.58134.5%
18. [15] 17.1038.3%17146.4%18.84139.3%9714.07124.6%
19. [14] 17.1038.3%17146.4%18.69138.2%10364.25132.9%
20. [08] 17.1038.3%17146.3%17.78131.5%10492.31134.5%
21. [13] 17.1038.3%17146.5%17.63130.4%9786.37125.5%
22. [05] 17.1038.3%17046.6%17.22127.4%8712.19111.7%
23. [09] 17.2038.5%17146.3%17.56129.9%9261.88118.8%
24. [01] 17.2038.5%17146.5%17.14126.8%8712.19111.7%
25. [02] 17.2038.5%17146.5%18.30135.4%10122.97129.8%
26. [06] 17.2038.5%17046.6%18.39136.0%10462.55134.2%
27. [12] 17.2038.5%17146.5%17.77131.4%10495.59134.6%
28. [03] 17.2038.5%17146.5%18.59137.5%9250.41118.6%
29. [11] 17.2038.5%17246.3%19.02140.7%9664.92123.9%
30. [07] 17.2038.5%17046.6%18.59137.5%9769.78125.3%
31. [04] 17.3038.7%17046.6%16.74123.8%8704.68111.6%
32. [10] 17.3038.7%17246.3%19.02140.7%10507.06134.7%
33. [00] 17.3038.7%17146.5%17.21127.3%9176.68117.7%

Impact of the option "Inlining":

# [# / #] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE Faster
in sec.
*Faster
in % *
Shorter
in KB
Shorter
in %
Slower
in sec.
Slower
in %
change
in KB
change
in %
1. [10/26] 4.0023.12%-5-2.93%4.1821.98%-2108620-20.07%
2. [03/19] 3.9022.67%-4-2.89%3.5919.31%-2681241-28.99%
3. [02/18] 3.9022.67%-4-2.89%4.1622.73%-1657943-16.38%
4. [11/27] 3.9022.67%-5-2.93%4.1521.82%-2137702-22.12%
5. [06/22] 3.9022.67%-4-2.88%3.4118.54%245760.23%
6. [14/30] 3.8022.22%-5-2.93%3.8620.65%-335189-3.23%
7. [07/23] 3.8022.09%-4-2.89%3.2917.70%-2146850-21.97%
8. [15/31]3.7021.64%-5-2.95%3.7619.96%-2201190-22.66%
9. [04/20] 3.4019.65%-4-2.85%2.3213.86%-1729467-19.87%
10. [09/25] 3.3019.19%-4-2.87%1.9911.33%-853606-9.22%
11. [12/28] 3.3019.19%-4-2.88%1.9811.14%966650.92%
12. [00/16] 3.3019.08%-4-2.85%1.7810.34%-1271398-13.85%
13. [13/29] 3.2018.71%-4-2.86%1.9911.29%-13721-0.14%
14. [05/21] 3.2018.71%-4-2.87%1.9611.38%-1010892-11.60%
15. [01/17] 3.2018.60%-4-2.88%1.659.63%-969386-11.13%
16. [08/24] 3.0017.54%-4-2.87%1.9110.74%57340.05%

Impact of the option "Moving while! and until!":

# [# / #] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE Faster
in sec.
*Faster
in % *
Shorter
in KB
Shorter
in %
Slower
in sec.
Slower
in %
change
in KB
change
in %
1. [00/08] 0.201.16%0-0.31%0.573.31%-1315635-14.34%
2. [17/25] 0.100.71%0-0.32%0.764.04%-433902-4.48%
3. [06/14] 0.100.58%0-0.33%0.301.63%983040.94%
4. [07/15] 0.100.58%0-0.33%0.251.34%557050.57%
5. [04/12] 0.100.58%0-0.31%1.036.15%-1790907-20.57%
6. [23/31]0.000.00%0-0.39%0.723.29%13650.01%
7. [03/11] 0.000.00%0-0.33%0.432.31%-414515-4.48%
8. [21/29] 0.000.00%0-0.32%0.442.29%-77004-0.79%
9. [19/27] 0.000.00%0-0.37%0.994.46%1290241.08%
10. [01/09] 0.000.00%0-0.33%0.422.45%-549683-6.31%
11. [22/30] 0.000.00%0-0.38%0.753.44%-261461-2.50%
12. [05/13] 0.000.00%0-0.33%0.412.38%-1074176-12.33%
13. [18/26] 0.000.00%0-0.38%0.743.29%-834764-7.09%
14. [20/28] 0.000.00%0-0.35%0.693.62%352250.34%
15. [02/10] -0.10-0.58%0-0.34%0.723.93%-384087-3.79%
16. [16/24] -0.10-0.71%0-0.33%0.703.69%-38502-0.37%

Impact of the option "Hoisting Iter Initialization":

# [# / #] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE Faster
in sec.
*Faster
in % *
Shorter
in KB
Shorter
in %
Slower
in sec.
Slower
in %
change
in KB
change
in %
1. [24/28] 0.201.42%00.25%0.060.30%876540.84%
2. [10/14] 0.201.16%00.30%-0.33-1.74%1428131.36%
3. [17/21] 0.100.71%00.27%0.392.08%-41506-0.43%
4. [16/20] 0.100.71%00.27%0.070.37%139260.13%
5. [09/13] 0.100.58%00.26%0.070.40%-524492-5.66%
6. [11/15] 0.100.58%00.29%-0.18-0.95%-49152-0.51%
7. [01/05] 0.100.58%00.26%0.080.47%00.00%
8. [25/29] 0.000.00%00.27%0.070.36%3153923.12%
9. [03/07] 0.000.00%00.29%0.000.00%-519372-5.61%
10. [26/30] 0.000.00%00.30%-0.65-2.80%191624515.19%
11. [18/22] 0.000.00%00.29%-0.66-2.94%134294111.40%
12. [02/06] 0.000.00%00.29%0.090.49%-339577-3.35%
13. [00/04] 0.000.00%00.26%-0.47-2.73%4719955.14%
14. [08/12] -0.10-0.58%00.26%-0.01-0.06%-3276-0.03%
15. [19/23] -0.10-0.75%00.29%-0.30-1.35%150180.13%
16. [27/31]-0.10-0.75%00.27%-0.57-2.46%-112640-0.95%

Impact of the option "Hoisting Loop Invariants":

# [# / #] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE Faster
in sec.
*Faster
in % *
Shorter
in KB
Shorter
in %
Slower
in sec.
Slower
in %
change
in KB
change
in %
1. [24/26] 0.805.67%0-0.13%3.5117.83%-2129100-20.30%
2. [17/19] 0.705.00%0-0.07%3.3918.04%-2250069-23.24%
3. [16/18] 0.705.00%0-0.08%3.4718.27%-1332838-12.76%
4. [20/22] 0.604.32%0-0.05%2.7414.38%-3822-0.04%
5. [28/30] 0.604.32%0-0.08%2.8014.18%-300509-2.89%
6. [25/27] 0.604.32%0-0.12%3.6218.52%-1687142-16.68%
7. [21/23] 0.503.60%0-0.05%2.7014.08%-2193544-22.56%
8. [29/31]0.503.60%0-0.12%2.9815.19%-2115174-21.58%
9. [12/14] 0.100.58%0-0.03%0.925.18%1313451.25%
10. [00/02] 0.100.58%0-0.05%1.096.33%-946293-10.31%
11. [04/06] 0.100.58%0-0.02%1.659.86%-1757866-20.19%
12. [13/15] 0.000.00%0-0.03%1.216.86%722940.74%
13. [09/11] 0.000.00%0-0.06%1.468.31%-403046-4.35%
14. [01/03] 0.000.00%0-0.07%1.458.46%-538214-6.18%
15. [05/07] -0.10-0.58%0-0.04%1.377.96%-1057587-12.14%
16. [08/10] -0.20-1.17%0-0.07%1.246.97%-14745-0.14%

Impact of the option "Common Subexpression Elimination":

# [# / #] Compiler Options *Best Run* Size of Executable Compile Time Memory Usage
Inline Move
while!
Once
Args.
Hoist
Invar.
CSE Faster
in sec.
*Faster
in % *
Shorter
in KB
Shorter
in %
Slower
in sec.
Slower
in %
change
in KB
change
in %
1. [24/25] 0.201.42%0-0.00%-0.14-0.71%3710973.54%
2. [04/05] 0.201.16%00.02%0.482.87%-7509-0.09%
3. [12/13] 0.100.58%0-0.00%-0.14-0.79%7092226.76%
4. [00/01] 0.100.58%00.02%-0.07-0.41%4644865.06%
5. [10/11] 0.100.58%00.01%0.000.00%8421378.01%
6. [14/15] 0.000.00%00.00%0.150.80%6501716.27%
7. [06/07] 0.000.00%00.00%0.201.09%6927706.62%
8. [26/27] 0.000.00%00.01%-0.03-0.13%8130566.44%
9. [18/19] 0.000.00%00.00%-0.28-1.25%-150732-1.28%
10. [02/03] 0.000.00%00.00%0.291.58%8725658.62%
11. [28/29] 0.000.00%00.01%-0.13-0.66%5988355.76%
12. [20/21] 0.000.00%0-0.01%0.120.63%7110656.81%
13. [16/17] 0.000.00%0-0.01%-0.20-1.05%7664987.34%
14. [08/09] -0.10-0.58%00.00%-0.22-1.24%123043811.73%
15. [22/23] -0.10-0.75%0-0.00%0.080.37%-1478656-14.17%
16. [30/31]-0.10-0.75%0-0.02%0.050.22%-1215829-11.36%