spectral-norm benchmark N=550

Each chart bar shows how many times slower, one ↓ spectral-norm program was, compared to the fastest program.

These are not the only programs that could be written. These are not the only compilers and interpreters. These are not the only programming languages.

Column × shows how many times more each program used compared to the benchmark program that used least.

    sort sortsort
  ×   Program Source Code CPU secs Elapsed secs Memory KB Code B ≈ CPU Load
1.0Python 2 #2 0.080.08?394  0% 0% 11% 100%
1.0Python 2 #2 0.080.08?394  0% 13% 0% 89%
1.6Python 3 #2 0.130.13?394  0% 0% 0% 100%
1.6PyPy 3 #8 0.130.13?594  20% 100% 0% 0%
1.6Python 3 #2 0.130.13?394  0% 0% 7% 100%
1.6PyPy 3 #8 0.130.13?594  0% 0% 100% 0%
1.6Python 3 #2 0.130.13?394  8% 0% 0% 100%
1.7PyPy 3 #8 0.130.13?594  93% 0% 0% 0%
1.7Python 2 #2 0.090.14?394  7% 29% 67% 7%
1.7PyPy 2 #8 0.140.14?594  36% 0% 0% 100%
1.7PyPy 2 #8 0.140.14?594  0% 100% 0% 0%
1.7PyPy 2 #8 0.140.14?594  7% 0% 0% 93%
2.9PyPy 2 #6 0.230.231,484498  8% 100% 0% 0%
2.9PyPy 2 #6 0.230.231,588498  9% 0% 0% 96%
2.9PyPy 2 #6 0.230.231,620498  8% 100% 8% 0%
3.5PyPy 3 #6 0.290.291,316498  7% 0% 0% 100%
3.6PyPy 3 #6 0.290.291,308498  7% 0% 0% 97%
3.6PyPy 3 #6 0.290.291,360498  7% 0% 100% 3%
4.8Python 3 #3 0.390.391,876642  5% 26% 76% 5%
4.8Python 3 #3 0.390.392,072642  100% 5% 3% 0%
4.9Python 3 #3 0.400.401,824642  8% 13% 100% 10%
5.0Nuitka #3 0.410.4138,304642  2% 0% 0% 100%
5.0Nuitka #3 0.410.4133,420642  2% 2% 100% 0%
5.1Nuitka #3 0.410.4138,168642  0% 0% 100% 0%
5.2Pyston #2 0.420.4256,036394  0% 0% 0% 100%
5.2Pyston #2 0.420.4255,652394  2% 0% 0% 100%
5.2Pyston #2 0.420.4257,532394  0% 10% 83% 12%
8.4PyPy 2 #5 1.290.68313,896595  54% 42% 41% 62%
8.4PyPy 2 #5 1.260.68311,756595  62% 37% 42% 51%
8.6PyPy 2 #5 1.290.69314,244595  46% 49% 60% 40%
11Numba 0.910.9184,284663  0% 0% 0% 100%
11Numba 0.920.9284,664663  0% 3% 49% 48%
11Numba 0.930.9383,916663  1% 0% 0% 99%
15PyPy 3 #5 2.201.1986,544575  38% 37% 43% 74%
15PyPy 3 #5 2.201.2086,116575  51% 41% 39% 64%
15PyPy 3 #5 2.221.2085,868575  41% 71% 44% 38%
18Pyston #5 4.751.43158,316595  86% 81% 85% 82%
18Pyston #5 4.791.49158,188595  88% 79% 76% 79%
19Pyston #5 4.821.50158,392595  88% 79% 76% 77%
21Pyston #8 1.691.6927,440594  100% 1% 0% 0%
21Pyston #8 1.691.6927,568594  0% 100% 0% 0%
21Pyston #8 1.731.7327,432594  1% 0% 2% 98%
22Nuitka #5 6.571.7759,500575  93% 97% 92% 93%
22Nuitka #5 6.551.7759,064575  95% 93% 92% 92%
23Python 2 #5 6.381.8441,668595  90% 93% 91% 92%
23Nuitka #5 6.631.8958,876575  91% 87% 89% 89%
23Python 2 #5 6.451.8941,868595  95% 91% 92% 93%
24Python 2 #5 6.341.9541,648595  92% 92% 91% 93%
27Python 3 #5 7.912.1654,288575  93% 92% 92% 91%
27Python 3 #5 7.992.1954,068575  94% 95% 93% 94%
27Pyston #6 2.202.2127,500498  0% 0% 100% 0%
27Pyston #6 2.212.2127,340498  2% 98% 0% 0%
27Python 3 #5 7.922.2355,120575  91% 90% 96% 90%
29Pyston #6 2.332.3327,464498  0% 100% 0% 0%
32Python development version #5 9.512.5649,964575  93% 93% 94% 94%
32Python development version #5 9.492.5752,260575  94% 92% 94% 94%
32Python development version #5 9.612.6150,292575  94% 93% 97% 95%
35Python 2 #6 2.802.806,720498  1% 3% 1% 100%
35Python 2 #6 2.802.806,664498  2% 2% 100% 2%
35Python 2 #6 2.812.826,724498  2% 1% 100% 1%
35Nuitka #8 2.862.8610,472594  1% 2% 0% 100%
35Nuitka #8 2.862.8710,376594  81% 1% 21% 1%
36Nuitka #8 2.892.9010,484594  4% 3% 100% 2%
38Nuitka #6 3.053.0510,300498  5% 5% 4% 100%
39Python 2 #8 3.123.126,864594  1% 2% 100% 1%
39Nuitka #6 3.123.1310,276498  12% 8% 96% 8%
39Python 2 #8 3.153.156,932594  2% 3% 1% 100%
39Nuitka #6 3.143.1510,296498  39% 10% 9% 71%
39Python 2 #8 3.163.166,796594  2% 1% 100% 1%
42Python 3 #6 3.363.379,168498  2% 61% 40% 1%
42Python 3 #6 3.383.399,404498  1% 1% 1% 100%
42Python 3 #6 3.383.399,096498  64% 1% 37% 1%
48Python development version #6 3.883.887,692498  2% 2% 100% 2%
49Python development version #6 3.943.947,484498  3% 1% 2% 100%
49Python development version #6 3.953.957,484498  12% 2% 90% 3%
51Python 3 #8 4.094.108,888594  74% 0% 27% 0%
51Python 3 #8 4.114.118,984594  25% 1% 77% 0%
51IronPython #8 3.984.1657,520594  0% 63% 8% 24%
52IronPython #8 3.994.1861,472594  92% 0% 1% 1%
52IronPython #8 4.004.1957,200594  0% 92% 1% 1%
53Python 3 #8 4.254.259,216594  1% 100% 2% 1%
57IronPython #6 4.344.5974,580498  1% 2% 13% 79%
57IronPython #6 4.354.6170,740498  7% 57% 1% 29%
57IronPython #6 4.454.6271,112498  46% 1% 48% 1%
59Jython #8 8.084.77294,356594  43% 34% 60% 31%
60Jython #8 8.374.87297,264594  51% 45% 38% 38%
61Jython #8 8.414.96294,976594  31% 64% 46% 29%
64Python development version #8 5.175.187,788594  2% 100% 2% 2%
65Python development version #8 5.275.277,756594  2% 1% 2% 99%
67Python development version #8 5.445.447,692594  55% 3% 47% 2%
84Jython #6 9.576.83284,740498  43% 37% 27% 33%
85Jython #6 9.826.89290,672498  30% 38% 31% 44%
85Jython #6 9.966.90294,640498  31% 36% 56% 21%
88MicroPython #6 7.127.154,424498  2% 100% 2% 3%
90MicroPython #6 7.267.294,156498  38% 100% 9% 7%
92MicroPython #6 7.437.474,452498  9% 78% 29% 13%
98Graal #6 10.317.93452,788498  24% 20% 19% 88%
98Graal #6 10.487.94447,692498  42% 32% 66% 26%
101Graal #6 10.468.15447,428498  84% 33% 32% 14%
missing benchmark programs
Cython No program
Shedskin No program
Grumpy No program

 spectral-norm benchmark : Eigenvalue using the power method

diff program output N = 100 with this output file to check your program is correct before contributing.

We are trying to show the performance of various programming language implementations - so we ask that contributed programs not only give the correct result, but also use the same algorithm to calculate that result.

Each program should calculate the spectral norm of an infinite matrix A, with entries a11=1, a12=1/2, a21=1/3, a13=1/4, a22=1/5, a31=1/6, etc

For more information see challenge #3 in Eric W. Weisstein, "Hundred-Dollar, Hundred-Digit Challenge Problems" and "Spectral Norm".

From MathWorld--A Wolfram Web Resource.
http://mathworld.wolfram.com/Hundred-DollarHundred-DigitChallengeProblems.html
http://mathworld.wolfram.com/SpectralNorm.html

Thanks to Sebastien Loisel for this benchmark.

Revised BSD license

  Home   Conclusions   License   Play