How big is the measured performance difference?
Each chart bar shows how many times more Time or how many times more Memory one unidentified ↓ fannkuch program used, compared to the program that used least Time or the program that used least Memory.
fannkuch benchmark N=12
This table shows 5 measurements - CPU Time, Elapsed Time, Memory, Code and ~ CPU Load.
Compare how much Memory the fannkuch 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 gcc #2 | 50.54 | 12.88 | 552 | 937 | 98% 100% 98% 98% |
| 1.1 | C++ GNU g++ #3 | 49.61 | 13.63 | 1,160 | 1186 | 81% 77% 100% 100% |
| 1.4 | Java 6 -server | 66.26 | 18.43 | 13,776 | 1150 | 78% 77% 98% 99% |
| 1.4 | Java 6 steady state | 67.43 | 18.59 | 14,608 | 1232 | 81% 77% 99% 99% |
| 2.4 | Ada 2005 GNAT #2 | 124.52 | 31.29 | 1,792 | 1518 | 99% 99% 100% 100% |
| 2.8 | Scala #3 | 132.57 | 36.01 | 105,476 | 1001 | 90% 91% 94% 91% |
| 3.1 | Haskell GHC #5 | 156.66 | 40.06 | 4,040 | 1204 | 96% 96% 96% 96% |
| 3.5 | C# Mono #2 | 162.98 | 44.48 | 5,584 | 1085 | 82% 79% 100% 100% |
| 3.6 | OCaml #2 | 178.79 | 46.77 | 4,004 | 975 | 93% 100% 94% 95% |
| 4.1 | Fortran Intel | 52.26 | 52.26 | 408 | 521 | 0% 100% 0% 0% |
| 4.6 | Pascal Free Pascal | 58.79 | 58.78 | 48 | 712 | 0% 100% 0% 0% |
| 4.8 | OCaml | 61.42 | 61.42 | 760 | 485 | 0% 100% 0% 0% |
| 4.9 | C GNU gcc | 63.73 | 63.73 | 488 | 628 | 0% 100% 0% 0% |
| 5.2 | ATS | 66.63 | 66.63 | 508 | 1191 | 100% 0% 0% 0% |
| 5.4 | Clean | 69.08 | 69.09 | 532 | 925 | 0% 100% 0% 0% |
| 5.5 | C++ GNU g++ #2 | 71.33 | 71.33 | 1,092 | 554 | 0% 0% 0% 100% |
| 5.6 | Lisp SBCL #2 | 71.69 | 71.69 | 5,100 | 624 | 0% 0% 0% 100% |
| 6.3 | Java 6 -server #4 | 80.88 | 80.75 | 13,144 | 555 | 0% 100% 0% 0% |
| 6.4 | Ada 2005 GNAT | 82.74 | 82.73 | 1,224 | 746 | 0% 0% 0% 100% |
| 8.4 | Scala #2 | 108.49 | 108.47 | 13,256 | 473 | 0% 0% 0% 100% |
| 8.7 | C# Mono | 112.21 | 112.20 | 5,044 | 549 | 0% 0% 0% 100% |
| 8.8 | Lua LuaJIT | 113.19 | 113.19 | 860 | 492 | 0% 0% 100% 0% |
| 9.4 | Go 6g 8g | 121.65 | 121.65 | 608 | 523 | 31% 0% 0% 69% |
| 10 | F# Mono #5 | 134.25 | 134.24 | 6,796 | 470 | 0% 0% 0% 100% |
| 14 | F# Mono #3 | 178.22 | 178.20 | 7,040 | 522 | 0% 0% 0% 100% |
| 14 | Java 6 -Xint | 11 min | 184.98 | 10,308 | 1150 | 81% 78% 100% 96% |
| 17 | F# Mono #4 | 222.26 | 222.35 | 7,036 | 497 | 77% 23% 0% 0% |
| 17 | Erlang HiPE #3 | 14 min | 223.03 | 6,276 | 864 | 95% 97% 98% 98% |
| 42 | Scheme PLT #2 | 9 min | 9 min | 9,176 | 549 | 0% 0% 100% 0% |
| 48 | Smalltalk VisualWorks | 10 min | 10 min | 12,724 | 772 | 0% 1% 0% 100% |
| 56 | JavaScript V8 #2 | 12 min | 12 min | 8,088 | 488 | 0% 100% 0% 0% |
| 66 | Java 6 -Xint #4 | 14 min | 14 min | 10,184 | 555 | 0% 0% 0% 100% |
| 83 | Perl #3 | 1h 10 min | 17 min | 6,020 | 471 | 99% 99% 99% 100% |
| 97 | PHP #3 | 1h 22 min | 20 min | 17,580 | 745 | 99% 100% 100% 100% |
| 111 | Haskell GHC #3 | 23 min | 23 min | 3,164 | 341 | 1% 99% 0% 0% |
| 126 | Scala | 27 min | 27 min | 169,452 | 453 | 13% 2% 6% 80% |
| 130 | JavaScript TraceMonkey #2 | 27 min | 27 min | 2,400 | 488 | 0% 0% 0% 100% |
| 146 | Ruby JRuby #3 | 31 min | 31 min | 92,472 | 340 | 12% 6% 34% 48% |
| 156 | Lua | 33 min | 33 min | 904 | 492 | 100% 0% 0% 0% |
| 321 | Perl #2 | 1h 08 min | 1h 08 min | 1,696 | 391 | 51% 35% 14% 0% |
| 323 | Mozart/Oz #2 | 1h 09 min | 1h 09 min | 4,564 | 506 | 0% 0% 100% 0% |
| 324 | Perl | 1h 09 min | 1h 09 min | 1,648 | 348 | 6% 94% 0% 0% |
| 338 | Python CPython #2 | 1h 12 min | 1h 12 min | 2,812 | 387 | 100% 0% 0% 0% |
| 379 | Python 3 #2 | 1h 21 min | 1h 21 min | 3,964 | 400 | 21% 0% 0% 79% |
| 390 | Ruby 1.9 #4 | 1h 15 min | 1h 23 min | 3,016 | 344 | 3% 97% 0% 0% |
| 526 | Ruby MRI #3 | 1h 52 min | 1h 52 min | 1,888 | 340 | 0% 100% 0% 0% |
| Erlang HiPE | Timed Out | 1h 00 min | 594 | |||
| Lisaac | Make Error | 788 | ||||
| Ruby 1.9 #3 | Bad Output | 340 | ||||
| interesting alternative programs | ||||||
| 1.9 | C# Mono #3 | 87.38 | 24.16 | 5,436 | 1089 | |
| 6.7 | Java 6 -server #3 | 86.21 | 86.20 | 12,112 | 914 | |
fannkuch benchmark : Indexed-access to tiny integer-sequence
diff program output N = 7 with this output file to check your program is correct before contributing.
Each program should
- "Take a permutation of {1,...,n}, for example: {4,2,1,5,3}.
- Take the first element, here 4, and reverse the order of the first 4 elements: {5,1,2,4,3}.
- Repeat this until the first element is a 1, so flipping won't change anything more: {3,4,2,1,5}, {2,4,3,1,5}, {4,2,3,1,5}, {1,3,2,4,5}.
- Count the number of flips, here 5.
- Do this for all n! permutations, and record the maximum number of flips needed for any permutation.
- Write the first 30 permutations and the number of flips.
The conjecture is that this maximum count is approximated by n*log(n) when n goes to infinity.
FANNKUCH is an abbreviation for the German word Pfannkuchen, or pancakes, in analogy to flipping pancakes."
The fannkuch benchmark is defined in Performing Lisp Analysis of the FANNKUCH Benchmark, Kenneth R. Anderson and Duane Rettig (26KB postscript)