diff options
Diffstat (limited to 'doc')
| -rw-r--r-- | doc/h48.md | 19 |
1 files changed, 11 insertions, 8 deletions
| @@ -251,15 +251,18 @@ Moreover, as an additional heuristic, in case of a 0 read we also look | |||
| 251 | up another pruning value in a table that takes into account only the | 251 | up another pruning value in a table that takes into account only the |
| 252 | position of the edges. This table is small (around 1MB), so repeated | 252 | position of the edges. This table is small (around 1MB), so repeated |
| 253 | accesses to it are not too slow. In practice, this gives a small speed up | 253 | accesses to it are not too slow. In practice, this gives a small speed up |
| 254 | of around 5%. More tables could be used to refine the fallback estimate, | 254 | of around 5% for random scrambles. However, for small solvers (low values |
| 255 | but each additional table leads to longer lookup times, especially if | 255 | of h) and positions where corners are close to solved, this fallback |
| 256 | it is too large to fit in cache. | 256 | table gives dramatic improvements (up to a factor of 1000x in some |
| 257 | manual tests). | ||
| 257 | 258 | ||
| 258 | Previous versions of this implementation (up to commit 6c42463, or | 259 | More tables could be used to refine the fallback estimate, but each |
| 259 | to version 0.2) also included the possibility of a table with 4 bits | 260 | additional table leads to longer lookup times, especially if it is too |
| 260 | per entry, at least for the `h0` case. Such tables did not require | 261 | large to fit in cache. Previous versions of this implementation (up |
| 261 | a fallback lookup, but due to their large size they were not less | 262 | to commit 6c42463, or to version 0.2) also included the possibility of |
| 262 | efficient. Therefore, they have been removed. | 263 | a table with 4 bits per entry, at least for the `h0` case. Such tables |
| 264 | did not require a fallback lookup, but due to their large size they were | ||
| 265 | not less efficient. Therefore, they have been removed. | ||
| 263 | 266 | ||
| 264 | ### Estimation refinements | 267 | ### Estimation refinements |
| 265 | 268 | ||
