diff options
| author | Sebastiano Tronto <sebastiano@tronto.net> | 2023-11-02 20:05:29 +0100 |
|---|---|---|
| committer | Sebastiano Tronto <sebastiano@tronto.net> | 2023-11-02 20:05:29 +0100 |
| commit | a61382e9dec22ea1ff3fea2d59d3c605c0aa2aac (patch) | |
| tree | b7b7a120439d45c780a3f664462bf29f56af5df2 /README.md | |
| parent | a2b8556e3cdc59bbb4a89c2779d2094db975eaa5 (diff) | |
| download | nissy-core-a61382e9dec22ea1ff3fea2d59d3c605c0aa2aac.tar.gz nissy-core-a61382e9dec22ea1ff3fea2d59d3c605c0aa2aac.zip | |
Optimized compose for avx2
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 6 |
1 files changed, 2 insertions, 4 deletions
| @@ -17,7 +17,8 @@ $ make test | |||
| 17 | 17 | ||
| 18 | * inline moves for avx2 | 18 | * inline moves for avx2 |
| 19 | * fix base get_ and set_ macros (constant arguments?) | 19 | * fix base get_ and set_ macros (constant arguments?) |
| 20 | * optimize things that use get_ and set_ | 20 | * optimize inverse for avx2 |
| 21 | * other things to optimize? | ||
| 21 | 22 | ||
| 22 | ### Documentation and interface | 23 | ### Documentation and interface |
| 23 | 24 | ||
| @@ -38,12 +39,9 @@ $ make test | |||
| 38 | 39 | ||
| 39 | ### Things I need to learn: | 40 | ### Things I need to learn: |
| 40 | 41 | ||
| 41 | * Use AVX2 instructions, in particular | ||
| 42 | [_mm256_shuffle_epi8](https://www.intel.com/content/www/us/en/docs/cpp-compiler/developer-guide-reference/2021-10/mm256-shuffle-epi8.html)) | ||
| 43 | * Inspect compiled assembly | 42 | * Inspect compiled assembly |
| 44 | * Use valgrind tool cachegrind and other profiling tools | 43 | * Use valgrind tool cachegrind and other profiling tools |
| 45 | 44 | ||
| 46 | |||
| 47 | ## Internal representation of the cube | 45 | ## Internal representation of the cube |
| 48 | 46 | ||
| 49 | The plan (TODO) is to have multiple implementations: some that | 47 | The plan (TODO) is to have multiple implementations: some that |