fasta benchmark ≈240MB N=25,000,000
Each chart bar shows how many times slower, one ↓ fasta 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 | sort | sort | ||||
| × | Program Source Code | CPU secs | Elapsed secs | Memory KB | Code B | ≈ CPU Load |
|---|---|---|---|---|---|---|
| 1.0 | Fortran Intel #4 | 3.18 | 3.19 | 260 | 1327 | 0% 0% 2% 100% |
| 1.2 | Haskell GHC #2 | 4.28 | 3.80 | 2,276 | 979 | 41% 62% 5% 4% |
| 1.3 | C++ g++ #2 | 4.10 | 4.11 | 620 | 1105 | 55% 0% 0% 46% |
| 1.3 | C++ g++ #3 | 4.23 | 4.23 | 872 | 1286 | 0% 0% 0% 100% |
| 1.3 | Ada 2005 GNAT #5 | 4.24 | 4.24 | 1,232 | 2186 | 0% 0% 100% 0% |
| 1.4 | Fortran Intel #3 | 4.54 | 4.55 | 524 | 1190 | 0% 0% 1% 100% |
| 1.5 | C++ g++ | 4.64 | 4.64 | 620 | 1033 | 100% 0% 0% 1% |
| 1.5 | C gcc #4 | 4.92 | 4.93 | 292 | 1221 | 0% 0% 100% 0% |
| 1.6 | Java 7 #4 | 5.07 | 4.99 | 14,976 | 1507 | 2% 0% 100% 1% |
| 1.6 | ATS | 5.20 | 5.20 | 288 | 2104 | 0% 0% 100% 1% |
| 2.0 | Fortran Intel | 6.34 | 6.35 | 252 | 1155 | 0% 0% 1% 100% |
| 2.0 | Scala #3 | 6.63 | 6.45 | 22,136 | 1053 | 2% 0% 100% 1% |
| 2.1 | C++ g++ #4 | 6.58 | 6.58 | 292 | 1266 | 1% 0% 100% 0% |
| 2.2 | C gcc | 6.90 | 6.91 | 292 | 1185 | 100% 0% 0% 0% |
| 2.4 | Java 7 #2 | 7.81 | 7.72 | 15,240 | 1240 | 0% 1% 1% 100% |
| 2.5 | Pascal Free Pascal #4 | 7.94 | 7.95 | 8 | 1112 | 0% 0% 0% 100% |
| 2.6 | Lisp SBCL #6 | 8.17 | 8.18 | 8,272 | 1751 | 0% 0% 0% 100% |
| 2.7 | C# Mono #2 | 8.46 | 8.46 | 15,788 | 1180 | 0% 0% 0% 100% |
| 2.8 | Ada 2005 GNAT | 9.04 | 9.04 | 1,232 | 1346 | 0% 0% 100% 1% |
| 3.1 | Dart | 9.84 | 9.85 | 288,788 | 1421 | 0% 95% 5% 0% |
| 3.2 | Scala | 10.42 | 10.22 | 22,244 | 1080 | 0% 1% 1% 100% |
| 3.6 | F# Mono | 11.34 | 11.35 | 16,572 | 978 | 100% 0% 1% 1% |
| 3.7 | OCaml #3 | 11.75 | 11.76 | 1,636 | 1042 | 0% 0% 0% 100% |
| 4.1 | Lisp SBCL #3 | 13.00 | 13.01 | 30,220 | 1579 | 0% 0% 0% 100% |
| 4.3 | Racket #3 | 13.80 | 13.80 | 17,244 | 1276 | 0% 0% 1% 100% |
| 4.4 | Clojure #5 | 22.42 | 14.17 | 81,136 | 1839 | 23% 88% 22% 27% |
| 4.8 | Lisp SBCL | 15.12 | 15.13 | 30,372 | 1419 | 0% 1% 0% 100% |
| 5.1 | Lisp SBCL #2 | 16.24 | 16.24 | 30,368 | 1617 | 0% 0% 0% 100% |
| 5.2 | Go | 16.68 | 16.68 | 1,012 | 1036 | 77% 0% 0% 23% |
| 16 | Erlang HiPE #2 | 49.74 | 49.75 | 8,788 | 1164 | 0% 100% 0% 1% |
| 16 | Racket | 50.96 | 50.96 | 16,084 | 1054 | 0% 0% 100% 0% |
| 23 | PHP #4 | 72.92 | 72.94 | 2,528 | 1110 | 0% 0% 100% 0% |
| 30 | Erlang HiPE | 95.10 | 95.12 | 8,656 | 1039 | 99% 0% 2% 0% |
| 37 | Perl | 116.40 | 116.45 | 99,960 | 838 | 0% 94% 0% 7% |
| 50 | PHP #3 | 157.07 | 160.25 | 2,520 | 1030 | 76% 0% 4% 20% |
| 56 | Ruby 2.0 #6 | 176.22 | 178.52 | 185,964 | 772 | 53% 30% 15% 4% |
| 67 | Ruby 2.0 #4 | 211.97 | 214.62 | 435,028 | 904 | 39% 30% 29% 4% |
| 73 | Ruby 2.0 #5 | 229.84 | 232.36 | 4,888 | 987 | 48% 27% 25% 2% |
| 77 | Ruby JRuby | 245.83 | 245.59 | 584,152 | 760 | 56% 8% 8% 30% |
| 78 | Python 3 #2 | 246.95 | 249.75 | 4,600 | 788 | 24% 2% 25% 51% |
| 79 | Python 3 | 249.98 | 252.91 | 4,612 | 792 | 33% 2% 39% 28% |
| 98 | Perl #4 | 5 min | 5 min | 2,280 | 934 | 1% 0% 100% 1% |
| 101 | Ruby 2.0 #2 | 5 min | 5 min | 248,980 | 732 | 11% 85% 5% 1% |
| 135 | PHP #2 | 7 min | 7 min | 2,520 | 1006 | 63% 4% 2% 32% |
| 136 | Perl #2 | 7 min | 7 min | 2,280 | 886 | 62% 0% 38% 0% |
| Haskell GHC #4 | Bad Output | 1413 | ||||
| Haskell GHC | Bad Output | 1421 | ||||
| OCaml #6 | Failed | 1161 | ||||
| "wrong" (different) algorithm / less comparable programs | ||||||
| 0.7 | Perl #5 | 2.30 | 2.30 | 15,308 | 1113 | |
| 0.7 | C gcc #2 | 2.37 | 2.38 | 420 | 1169 | |
| 3.6 | Haskell GHC #3 | 12.84 | 11.41 | 1,748 | 1408 | |
fasta benchmark : Generate and write random DNA sequences
diff program output N = 1000 with this 10KB 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
- generate DNA sequences, by copying from a given sequence
- generate DNA sequences, by weighted random selection from 2 alphabets
- convert the expected probability of selecting each nucleotide into cumulative probabilities
- match a random number against those cumulative probabilities to select each nucleotide (use linear search or binary search)
- use this linear congruential generator to calculate a random number each time a nucleotide needs to be selected (don't cache the random number sequence)
IM = 139968 IA = 3877 IC = 29573 Seed = 42 Random (Max) Seed = (Seed * IA + IC) modulo IM = Max * Seed / IM
- write 3 sequences line-by-line in FASTA format
We'll use the generated FASTA file as input for other benchmarks (reverse-complement, k-nucleotide).
Random DNA sequences can be based on a variety of Random Models (554KB pdf). You can use Markov chains or independently distributed nucleotides to generate random DNA sequences online.