fannkuch-redux benchmark | Computer Language Benchmarks Game

fannkuch-redux benchmark N=12

Each chart bar shows how many times slower, one ↓ fannkuch-redux 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.

×	Program Source Code	CPU secs	Elapsed secs	Memory KB	Code B	≈ CPU Load
		sort		sort	sort
1.0	C++ g++ #7	22.53	22.54	372	1150	0% 0% 0% 100%
1.0	C gcc #4	22.69	22.70	372	1183	0% 0% 0% 100%
1.7	Ada 2005 GNAT #3	38.17	38.20	1,960	2100	0% 0% 0% 100%
2.0	C# Mono #2	44.82	44.83	13,844	564	0% 0% 0% 100%
2.1	C++ g++ #5	47.93	47.95	1,240	1440	0% 0% 0% 100%
2.1	C gcc #2	48.41	48.43	412	1557	0% 0% 0% 100%
2.2	Fortran Intel #3	48.97	48.99	7,596	1148	0% 0% 0% 100%
2.2	C gcc #3	50.35	50.37	372	567	0% 0% 0% 100%
2.3	ATS #2	51.46	51.48	360	1571	0% 0% 0% 100%
2.3	ATS	51.72	51.73	364	1306	0% 0% 0% 100%
2.3	Lisp SBCL #4	52.62	52.64	21,204	1518	0% 0% 0% 100%
2.4	C++ g++ #3	53.38	53.39	372	593	0% 0% 0% 100%
2.4	C++ g++	54.82	54.84	1,140	1059	0% 0% 0% 100%
2.6	Fortran Intel	57.62	57.64	244	590	0% 0% 0% 100%
2.7	C++ g++ #4	60.10	60.12	1,240	1439	0% 0% 0% 100%
2.7	Lisp SBCL #3	60.93	60.95	20,632	821	0% 0% 0% 100%
2.7	Scala #2	61.36	61.38	25,268	1017	0% 0% 0% 100%
2.8	F# Mono #3	62.14	62.16	20,428	945	0% 0% 0% 100%
2.9	OCaml #4	0.00	65.32	11,296	1004	0% 0% 0% 100%
2.9	OCaml #3	0.00	65.48	21,008	1017	0% 0% 0% 100%
2.9	Haskell GHC #4	66.34	66.36	1,972	658	0% 0% 0% 100%
3.1	Java 7	70.18	70.20	17,168	1282	0% 0% 0% 100%
3.2	C gcc	71.59	71.61	376	508	0% 0% 0% 100%
3.2	C# Mono #3	72.41	72.43	15,060	1096	0% 0% 0% 100%
3.3	Go	73.61	73.64	1,276	900	0% 0% 0% 100%
3.3	Pascal Free Pascal	75.35	75.37	824	1018	0% 0% 0% 100%
3.4	Java 7 #2	76.62	76.64	17,560	514	0% 0% 0% 100%
3.5	JavaScript V8 #3	78.67	78.70	6,312	539	0% 0% 0% 100%
3.5	C# Mono	79.41	79.43	13,844	520	0% 0% 0% 100%
3.6	F# Mono #2	80.07	80.09	17,004	548	0% 0% 0% 100%
3.7	Dart #2	82.44	82.46	8,988	538	0% 0% 0% 100%
3.8	JavaScript V8 #2	84.82	84.84	6,272	472	0% 0% 0% 100%
3.8	Lisp SBCL #2	86.70	86.72	4,780	513	0% 0% 0% 100%
4.1	OCaml #2	91.99	92.01	748	473	0% 0% 0% 100%
4.2	JavaScript V8	94.74	94.78	6,292	463	0% 0% 0% 100%
4.2	Scala	94.96	94.99	22,952	459	0% 0% 0% 100%
4.7	OCaml	106.85	106.87	752	524	0% 0% 0% 100%
5.9	Clojure #2	132.60	132.69	53,624	1088	0% 0% 0% 100%
13	Haskell GHC #2	288.46	288.56	3,376	808	0% 0% 0% 100%
14	Racket	5 min	5 min	18,008	649	0% 0% 0% 100%
14	F# Mono	5 min	5 min	15,004	551	0% 0% 0% 100%
20	Erlang HiPE	7 min	7 min	11,624	1038	0% 0% 0% 100%
26	Haskell GHC	9 min	9 min	3,360	553	0% 0% 0% 100%
31	Smalltalk VisualWorks	11 min	11 min	42,580	838	0% 0% 0% 100%
36	Ruby JRuby	13 min	13 min	599,808	384	0% 0% 0% 100%
71	Dart	26 min	26 min	8,660	531	0% 0% 0% 100%
94	Lua	35 min	35 min	1,016	462	0% 0% 0% 100%
102	Perl	38 min	38 min	1,836	457	0% 0% 0% 100%
104	Python 3 #6	39 min	39 min	6,152	385	0% 0% 0% 100%
134	PHP #3	50 min	50 min	14,716	1150	0% 0% 0% 100%
138	PHP #2	51 min	51 min	3,188	441	0% 0% 0% 100%
148	Ruby 2.0	55 min	55 min	6,148	384	0% 0% 0% 100%
156	PHP	58 min	58 min	3,232	482	0% 0% 0% 100%
	Haskell GHC #3	Timed Out	1h 00 min		1153
	Python 3	Timed Out	1h 00 min		1108
	Python 3 #2	Timed Out	1h 00 min		797
"wrong" (different) algorithm / less comparable programs
1.4	C++ g++ #2	31.52	31.54	1,196	2106
1.8	C++ g++ #6	40.52	40.54	376	894
2.0	Lisp SBCL	44.14	44.15	26,848	1607
2.1	Java 7 #3	47.96	47.98	19,468	1633

fannkuch-redux benchmark : Indexed-access to tiny integer-sequence

diff program output N = 7 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.

For N = 7 programs should generate these permutations (40KB) - which, incidentally, seem to be in the same order as permutations generated by the Tompkins-Paige algorithm, see pages 150-151 Permutation Generation Methods Robert Sedgewick.

The fannkuch benchmark is defined by programs in Performing Lisp Analysis of the FANNKUCH Benchmark, Kenneth R. Anderson and Duane Rettig.

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.
Keep a checksum
- checksum = checksum + (if permutation_index is even then flips_count else -flips_count)
- checksum = checksum + (toggle_sign_-1_1 * flips_count)
Do this for all n! permutations, and record the maximum number of flips needed for any permutation.

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.

Thanks to Oleg Mazurov for insisting on a checksum and providing this helpful description of the approach he took -

A common idea for parallel implementation is to divide all work (n! permutations) into chunks small enough to avoid load imbalance but large enough to keep overhead low. I set the number of chunks as a parameter (NCHUNKS = 150) from which I derive the size of a chunk (CHUNKSZ) and the actual number of chunks/tasks to be processed (NTASKS), which may be different from NCHUNKS because of rounding.
Task scheduling is trivial: threads will atomically get and increment the taskId variable to derive a range of permutation indices to work on:
```
    task = taskId.getAndIncrement();
    idxMin = task * CHUNKSZ;
    idxMax = min( idxMin + CHUNKSZ, n! );
```
Maximum flip counts and partial checksums can be computed for chunks in arbitrary order and recombined to generate the required result at the final step (CHUNKSZ must be even for adding partial checksums to be associative - I didn't enforce it in my submission).
Now I need to go from a permutation index to the permutation itself.
The predefined order in which all permutations are to be generated can be described as follows: to generate n! permutations of n arbitrary numbers, rotate the numbers left (from higher position to lower) n times, so that each number appears in the n-th position, and for each rotation recursively generate (n-1)! permutations of the first n-1 numbers whatever they are.
To optimize the process I use an intermediate data structure, count[], which keeps count of how many rotations have been done at every level. Apparently, count[0] is always 0, as there is only one element at that level, which can't be rotated; count[1] = 0..1 for two elements, count[2] = 0..2 for three elements, etc.
To generate next permutation I swap the first two elements and increase count[1]. If count[1] becomes greater than 1, I'm done with rotations at level 1 and need to "return" (as it would have been in the recursive implementation) to level 2. Now, I rotate 3 elements and increment count[2]. If it becomes greater than 2, I'm done with level 2 and need to go to level 3, etc.
It should be clear now how to generate a permutation and corresponding count[] array from an arbitrary index. Basically, count[k] = ( index % (k+1)! ) / k! is the number of rotations we need to perform on elements 0..k. Doing it in the descending order from n-1 to 1 gives us both the count[] array and the permutation.

Home Conclusions License Play

x64 Ubuntu : Intel® Q6600® one coreComputer Language Benchmarks Game [[ Play ]]

fannkuch-redux benchmark N=12

fannkuch-redux benchmark : Indexed-access to tiny integer-sequence

x64 Ubuntu : Intel® Q6600® one core
Computer Language Benchmarks Game [[ Play ]]