spectral-norm benchmark N=5,500

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
1.0Fortran Intel 7.847.851,348568  0% 0% 0% 100%
1.0Fortran Intel #2 7.847.85788513  1% 0% 0% 100%
1.0C gcc #5 7.857.85752569  1% 1% 1% 100%
1.0Fortran Intel #3 7.857.861,324638  0% 0% 1% 100%
1.0Rust #3 7.867.865,1161020  1% 1% 1% 100%
1.0Ada 2005 GNAT #4 7.877.881,9282762  0% 1% 1% 100%
1.3C++ g++ #5 10.0510.057481044  1% 1% 1% 100%
1.3C++ g++ #6 10.0510.057521044  1% 0% 1% 100%
1.4C gcc #4 10.6110.617521139  1% 0% 0% 100%
1.4C++ g++ #8 10.6310.647521278  1% 0% 1% 100%
2.0OCaml #2 15.6815.691,808377  0% 0% 0% 100%
2.0Haskell GHC #4 15.6915.701,588984  1% 0% 0% 100%
2.0Rust #2 15.6915.704,896485  1% 0% 0% 100%
2.0Go 15.7015.711,296411  1% 0% 1% 100%
2.0JavaScript V8 #3 15.7115.7110,532373  0% 1% 5% 100%
2.0JavaScript V8 #2 15.7115.7110,536328  1% 0% 1% 100%
2.0C gcc #3 15.7115.71752463  1% 1% 0% 100%
2.0Go #2 15.7115.711,300668  1% 0% 0% 100%
2.0JavaScript V8 15.7115.7110,540311  1% 0% 0% 100%
2.0Go #3 15.7115.721,556536  1% 0% 1% 100%
2.0Pascal Free Pascal 15.7115.728423  1% 1% 0% 100%
2.0Lisp SBCL #2 15.8715.8823,060906  0% 1% 1% 100%
2.0Lisp SBCL #3 15.8715.8823,068883  0% 1% 1% 100%
2.0Dart 15.8815.8916,900457  1% 0% 0% 100%
2.1Ada 2005 GNAT #3 16.5016.521,9281702  1% 2% 1% 100%
2.1Rust 16.6016.6237,860811  1% 1% 1% 100%
2.1OCaml #3 15.9216.723,556938  0% 0% 0% 100%
2.1Scala 16.7616.7627,780404  1% 0% 0% 100%
2.2Java  17.0917.1025,552514  1% 0% 0% 100%
2.2Java  #2 17.4117.4125,568950  0% 1% 1% 100%
2.3Clojure #8 17.9417.9561,864918  0% 1% 1% 100%
2.3Clojure #6 18.3118.3261,732808  1% 1% 0% 100%
2.4Clojure #7 18.5018.5261,248762  1% 0% 1% 100%
2.6Racket #3 20.6020.6126,176627  0% 1% 0% 100%
2.6C gcc 20.6920.70468383  1% 1% 0% 100%
2.6C++ g++ 20.7520.761,072452  1% 1% 0% 100%
2.7Lisp SBCL 20.9220.9212,720625  0% 1% 0% 100%
2.8Ada 2005 GNAT 21.7621.771,492710  1% 0% 1% 100%
2.8C# Mono 21.8621.8620,156459  1% 0% 0% 100%
2.8C# Mono #2 21.8821.8922,3121063  0% 1% 1% 100%
2.9Haskell GHC #2 22.7622.774,036403  0% 1% 0% 100%
3.0F# Mono #2 22.2123.1930,012852  0% 1% 0% 96%
4.8Racket #2 37.7837.8023,372532  0% 1% 1% 100%
5.1Racket 40.1740.1819,868446  0% 0% 1% 100%
7.1Hack #2 55.5755.5984,712398  0% 0% 1% 100%
7.1Hack #3 55.9355.96162,5121195  1% 1% 1% 100%
8.7Erlang HiPE #2 68.4768.5018,480747  1% 0% 0% 100%
12Smalltalk VisualWorks #2 91.0391.0542,880438  0% 0% 0% 100%
14Erlang HiPE 112.65112.6919,624507  1% 1% 0% 100%
25Perl 192.71192.785,372333  0% 0% 0% 100%
32Ruby JRuby #4 249.37249.79677,572326  0% 1% 1% 100%
32Lua 252.65252.721,288329  0% 0% 0% 100%
34Ruby JRuby 266.87267.33680,488292  0% 1% 1% 100%
36Ruby JRuby #2 279.66280.12681,104776  0% 1% 0% 100%
43Ruby #4 5 min5 min9,000326  0% 1% 0% 100%
49Ruby 6 min6 min9,108292  1% 2% 1% 100%
51Ruby #3 6 min6 min42,020828  1% 1% 0% 100%
55PHP #2 7 min7 min7,240397  0% 0% 0% 100%
58PHP #3 7 min7 min24,4361193  0% 0% 0% 100%
95Python 3 #6 12 min12 min6,468328  1% 0% 0% 100%
106Python 3 #5 13 min13 min43,632437  0% 1% 1% 100%
114Perl #3 14 min14 min9,592846  0% 0% 0% 100%
115Perl #2 15 min15 min3,816343  0% 0% 0% 100%
124Python 3 #8 16 min16 min5,600449  1% 0% 0% 100%
Perl #4 Failed551
Scala #4 Failed1006
Scala #3 Failed982
Scala #2 Failed720
"wrong" (different) algorithm / less comparable programs
0.3C gcc #2 2.332.33236,908669
1.0C++ g++ #2 7.857.867521330
1.1C++ g++ #7 8.428.437521283
1.4Python 3 #2 11.0411.05488,572233
12Python 3 #3 96.1796.751,198,228379

 spectral-norm benchmark : Eigenvalue using the power method

You can write your own program for this task and contribute to the benchmarks game by following these general instructions.

More specifically:

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

Each program must implement 4 separate functions / procedures / methods like the C# program.

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