Benchmark examples

This page explains what kind of problems we use for our benchmarking.

The column Solutions states whether search was performed for one, all, or the best solution. The column S/F/C counts the solutions, failures, and choices in the problem's search tree. The link to a problem's source code points into our documentation.

Click on the name of an example to get further information, like propagators and branchings used, and the command line used to execute this example with the given options.

Example Solutions S/F/C Link to source code
Alpha (smart) All 1/33/33 alpha.cc
Alpha (naive) All 1/7435/7435 alpha.cc
BIBD 7,3,60 1 1/1306/1318 bibd.cc
18 Knights 1 1/12/266 knights.cc
Golomb rulers 10 Best 10/8892/8911 golomb-ruler.cc
10 Queens (naive) All 724/4992/5715 queens.cc
10 Queens (smart) All 724/4992/5715 queens.cc
100 Queens (naive) 1 1/22/115 queens.cc
100 Queens (smart) 1 1/22/115 queens.cc
Eq 20 All 1/54/54 eq20.cc
Graph coloring 1 1/1078/1208 graph-color.cc
Magic sequence 500 (naive) All 1/250/250 magic-sequence.cc
Magic sequence 500 (smart) All 1/250/250 magic-sequence.cc
Magic sequence 500 (using gcc) All 1/250/250 magic-sequence.cc
Photo alignment Best 5/10345/10354 photo.cc
Partition 32 1 1/160258/160297 examples_2partition.cc
Social golfers 9,8 1 1/32/215 golf.cc
Crew allocation 1 1/34/86 crew.cc
Steiner triples 1 1/1067/1091 steiner.cc
Domain stress All 0/1/0 stress-domain.cc
Failure stress All 0/1/0 stress-exec.cc
Search stress All 117649/0/117648 stress-search.cc