How big is the measured performance difference?
Each chart bar shows how many times more Time or how many times more Memory one unidentified ↓ spectral-norm program used, compared to the program that used least Time or the program that used least Memory.
spectral-norm benchmark N=5,500
This table shows 5 measurements - CPU Time, Elapsed Time, Memory, Code and ~ CPU Load.
Compare how much Memory the spectral-norm programs used - sort Memory KB. Compare how much Code the programs used - sort Code B
Column × shows how many times more each program used compared to the program that used least.
| sort | sort | sort | sort | |||
|---|---|---|---|---|---|---|
| × | Program Source Code | CPU secs | Elapsed secs | Memory KB | Code B | ~ CPU Load |
| 1.0 | C++ GNU g++ #7 | 11.98 | 3.00 | 1,032 | 1114 | 100% 100% 100% 100% |
| 1.0 | C GNU gcc #4 | 12.00 | 3.00 | 688 | 1139 | 100% 100% 100% 100% |
| 1.3 | Java 6 steady state #2 | 15.71 | 3.96 | 18,112 | 1027 | 99% 99% 99% 99% |
| 1.3 | C# Mono #2 | 15.76 | 3.99 | 4,712 | 1063 | 98% 99% 99% 99% |
| 1.3 | Fortran Intel | 15.82 | 4.00 | 1,140 | 568 | 99% 99% 99% 99% |
| 1.3 | Haskell GHC | 15.67 | 4.04 | 1,752 | 869 | 96% 98% 97% 98% |
| 1.4 | Java 6 -server #2 | 15.95 | 4.11 | 12,552 | 950 | 97% 97% 98% 97% |
| 1.5 | Scala #2 | 17.23 | 4.47 | 16,760 | 720 | 96% 98% 95% 96% |
| 1.5 | Lisp SBCL #3 | 15.61 | 4.53 | 5,728 | 883 | 86% 86% 86% 85% |
| 1.6 | Ada 2005 GNAT #2 | 18.74 | 4.74 | 2,744 | 1464 | 99% 99% 99% 99% |
| 1.6 | Ada 2005 GNAT #3 | 18.91 | 4.81 | 2,260 | 1702 | 98% 98% 98% 98% |
| 1.8 | ATS #2 | 20.57 | 5.54 | 1,268 | 2339 | 93% 93% 92% 92% |
| 1.9 | OCaml #3 | 18.58 | 5.80 | 3,112 | 907 | 77% 83% 79% 81% |
| 3.8 | Go 6g 8g #2 | 44.78 | 11.47 | 1,296 | 545 | 97% 97% 98% 98% |
| 3.9 | Erlang HiPE #2 | 46.05 | 11.82 | 10,552 | 747 | 97% 97% 98% 97% |
| 5.2 | Lua LuaJIT #3 | 15.69 | 15.70 | 1,072 | 328 | 0% 100% 0% 0% |
| 5.2 | Haskell GHC #3 | 15.70 | 15.70 | 1,472 | 737 | 0% 0% 100% 0% |
| 5.2 | Lua LuaJIT | 15.71 | 15.71 | 1,072 | 329 | 100% 0% 0% 0% |
| 5.2 | Lisp SBCL | 15.75 | 15.74 | 6,844 | 625 | 0% 0% 100% 0% |
| 5.3 | Fortran Intel #2 | 15.90 | 15.90 | 588 | 513 | 0% 0% 0% 100% |
| 5.7 | Scala | 17.45 | 17.15 | 13,880 | 500 | 0% 0% 100% 2% |
| 5.7 | Java 6 -server | 17.45 | 17.15 | 13,324 | 514 | 0% 100% 0% 2% |
| 5.9 | C# Mono | 17.72 | 17.72 | 4,376 | 459 | 0% 0% 100% 0% |
| 5.9 | F# Mono | 17.83 | 17.82 | 6,524 | 514 | 0% 0% 0% 100% |
| 6.2 | Clean | 18.73 | 18.73 | 1,440 | 572 | 0% 0% 0% 100% |
| 6.3 | Pascal Free Pascal | 18.79 | 18.79 | 184 | 423 | 0% 100% 0% 0% |
| 6.3 | OCaml #2 | 18.84 | 18.84 | 1,628 | 377 | 100% 0% 0% 0% |
| 6.9 | ATS | 20.66 | 20.66 | 584 | 1179 | 100% 0% 0% 0% |
| 7.3 | Ada 2005 GNAT | 21.77 | 21.76 | 1,336 | 710 | 0% 0% 0% 100% |
| 9.8 | Java 6 -Xint #2 | 117.30 | 29.54 | 10,840 | 950 | 99% 99% 99% 99% |
| 11 | C GNU gcc | 32.78 | 32.78 | 616 | 382 | 100% 0% 0% 0% |
| 11 | C++ GNU g++ | 32.79 | 32.79 | 1,144 | 452 | 0% 0% 0% 100% |
| 13 | Go 6g 8g | 39.38 | 39.38 | 1,444 | 420 | 0% 0% 0% 100% |
| 20 | JavaScript TraceMonkey | 58.58 | 58.59 | 4,072 | 311 | 0% 0% 100% 0% |
| 22 | Erlang HiPE | 66.11 | 66.11 | 8,804 | 507 | 0% 0% 100% 0% |
| 27 | JavaScript V8 | 81.21 | 81.20 | 13,880 | 311 | 100% 0% 0% 0% |
| 30 | Haskell GHC #2 | 89.75 | 88.69 | 8,952 | 396 | 1% 8% 93% 3% |
| 37 | Smalltalk VisualWorks #2 | 111.96 | 111.95 | 14,524 | 438 | 0% 0% 0% 100% |
| 49 | Java 6 -Xint | 148.51 | 148.49 | 11,484 | 514 | 0% 0% 100% 0% |
| 87 | Lua | 262.23 | 262.22 | 1,132 | 329 | 0% 0% 100% 0% |
| 88 | Lua #3 | 264.23 | 264.21 | 1,132 | 328 | 100% 0% 0% 0% |
| 88 | Scheme PLT #2 | 264.46 | 264.46 | 9,248 | 419 | 0% 0% 0% 100% |
| 99 | PHP #3 | 19 min | 296.76 | 16,624 | 1193 | 99% 98% 99% 98% |
| 99 | Python CPython #5 | 19 min | 296.88 | 20,744 | 428 | 95% 96% 97% 100% |
| 102 | Perl #3 | 20 min | 5 min | 7,792 | 846 | 100% 100% 99% 100% |
| 117 | Python 3 #5 | 23 min | 5 min | 25,384 | 437 | 98% 98% 99% 99% |
| 172 | Mozart/Oz | 8 min | 8 min | 3,380 | 411 | 100% 0% 0% 0% |
| 299 | Ruby JRuby | 14 min | 14 min | 95,456 | 292 | 0% 18% 82% 0% |
| 303 | Ruby 1.9 | 15 min | 15 min | 3,920 | 292 | 46% 54% 0% 0% |
| 385 | PHP #2 | 19 min | 19 min | 5,312 | 397 | 100% 0% 0% 0% |
| 389 | Perl #2 | 19 min | 19 min | 3,148 | 343 | 100% 0% 0% 0% |
| 390 | Python CPython | 19 min | 19 min | 3,384 | 378 | 99% 0% 0% 1% |
| 440 | Python 3 | 22 min | 22 min | 4,452 | 366 | 0% 100% 0% 0% |
| 465 | Perl | 23 min | 23 min | 2,812 | 355 | 0% 0% 0% 100% |
| 750 | Ruby MRI | 37 min | 37 min | 2,676 | 292 | 0% 0% 0% 100% |
| Lisaac | Make Error | 518 | ||||
| interesting alternative programs | ||||||
| 1.4 | C GNU gcc #2 | 4.18 | 4.19 | 236,972 | 669 | |
| 4.2 | Python CPython #2 | 12.71 | 12.72 | 479,908 | 233 | |
| 303 | Python CPython #4 | 910.43 | 910.38 | 3,452 | 281 | |
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.
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.