data-oriented-design-guide
/data-oriented-design-guide
Pomaga projektować i refaktoryzować układy danych krytyczne dla wydajności pod kątem pamięci podręcznej: SoA, AoS, podział pól gorących i zimnych, unikanie śledzenia wskaźników.
- Autor
- xonovex ↗
- Repozytorium
- platform ↗
- Licencja
- MIT
- Ostatnia zmiana repozytorium
- 08.10.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 16
- Skrypty
- 0
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 0
Brak sygnałów. Repozytorium zmienione 08.10.2026, bez odwołań do wycofanych modeli i API.
MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.
Co robi i kiedy
Pomaga projektować i refaktoryzować układy danych krytyczne dla wydajności pod kątem pamięci podręcznej: SoA, AoS, podział pól gorących i zimnych, unikanie śledzenia wskaźników.
Wyzwalacze z opisu
Oryginalny opis (EN)
Use when designing or refactoring performance-critical data layouts for cache efficiency, in any language. Triggers on prompts about SoA/AoS/AoSoA, cache misses, hot/cold field splitting, data-oriented design, component-storage layout, batch/bulk processing, handles vs pointers, avoiding pointer chasing, even when the user doesn't say 'data-oriented' or 'cache'.
Jak działa
- Essentials
- Layout
- Processing
- Memory
- Gotchas
- Progressive Disclosure
Podgląd SKILL.md
--- name: data-oriented-design-guide description: "Use when designing or refactoring performance-critical data layouts for cache efficiency, in any language. Triggers on prompts about SoA/AoS/AoSoA, cache misses, hot/cold field splitting, data-oriented design, component-storage layout, batch/bulk processing, handles vs pointers, avoiding pointer chasing, even when the user doesn't say 'data-oriented' or 'cache'." --- # Data-Oriented Design Guidelines Organize data around access patterns and measure cache traffic before changing performance-critical layouts. ## Essentials - **The problem is data movement** - Optimize cache traffic, not instruction count, see [references/cache-behavior.md](references/cache-behavior.md) - **Design from the data** - Model bulk input→output transforms, not idealized objects, see [references/data-as-transforms.md](references/data-as-transforms.md) - **Measure first** - Profile cache misses before changing layout, see [references/measurement-and-profiling.md](references/measurement-and-profiling.md) - **Record cheaply, always on** - Per-frame accumulating counters recorded via a cached accumulator pointer (one store, no lookup/branch/lock), history allocated only for viewed counters, see [references/statistics-recording.md](references/statistics-recording.md) ## Layout - **SoA / AoS / AoSoA** - Split structs into per-field arrays for hot loops, see [references/soa-aos-aosoa.md](references/soa-aos-aosoa.md) - **SIMD-friendly layout** - Contiguous columns, padding, alignment for vectorization, see [references/simd-friendly-layout.md](references/simd-friendly-layout.md) - **Nested / variable-length arrays** - Hold child lists in bulk arrays + cache-line chunks, not per-object heap pointers, see [references/nested-arrays.md](references/nested-arrays.md) ## Processing - **Sequential access** - Stream linearly, never chase pointers, see [references/access-patterns.md](references/access-patterns.md) - **Existence-based processing** - Bucket by state so branches become loops, see [references/existence-based-processing.md](references/existence-based-processing.md) - **Batch over one-at-a-time** - Transform N items per call, amortize overhead, see [references/data-as-transforms.md](references/data-as-transforms.md) - **Default to zero** - Make all-zero a valid default; reserve 0 for none/neutral, not magic sentinels, see [references/zero-as-default.md](references/zero-as-default.md) ## Memory - **Handles, not pointers** - Reference by index/handle into relocatable, stable arrays (swap-remove, free lists, indirection tables), see [references/handles-and-indices.md](references/handles-and-indices.md) - **Allocation** - Contiguous storage matters for cache, but the allocators themselves (arenas/pools/lifetimes) are general, see **memory-management-guide** ## Gotchas - SoA only wins when loops touch a subset of fields; full-record access can favor AoS: measure both. - The hardware prefetcher tracks linear strides; randomizing your index order silently disables it. - Padding for alignment trades memory for throughput, on cache-bound loads, the smaller packed layout can still win. - A counter you cache as a raw pointer dangles if its backing array reallocates: hand out indices into a stable block, or pointers into a non-relocating pool. ## Progressive Disclosure - Read [references/cache-behavior.md](references/cache-behavior.md) - Load when reasoning about cache lines, latency, or miss types - Read [references/data-as-transforms.md](references/data-as-transforms.md) - Load when modeling a system as bulk streams instead of objects - Read [references/soa-aos-aosoa.md](references/soa-aos-aosoa.md) - Load when choosing or converting between AoS, SoA, and AoSoA layouts - Read [references/access-patterns.md](references/access-patterns.md) - Load when iterating collections or replacing pointer-chasing structures - Read [references/handles-and-indices.md](references/handles-and-indices.md) - Load when designing entity references, stable arrays, or free lists - Read [references/existence-based-processing.md](references/existence-based-processing.md) - Load when removing per-item branches by sorting or bucketing - Read [references/simd-friendly-layout.md](references/simd-friendly-layout.md) - Load when laying out data for vectorization or alignment - Read [references/nested-arrays.md](references/nested-arrays.md) - Load when objects own variable-length child lists, or you're reaching for a per-object growable container - Read [references/zero-as-default.md](references/zero-as-default.md) - Load when choosing sentinels, nil values, or default-initialized state - Read [references/measurement-and-profiling.md](references/measurement-and-profiling.md) - Load when measuring before/after or reading hardware counters - Read [references/statistics-recording.md](references/statistics-recording.md) - Load when building always-on counters/telemetry that must add near-zero overhead on hot paths
Pliki w paczce (16)
- SKILL.md 5 KB
- SOURCES.md 4 KB
- eval-queries.json 5 KB
- evals.json 4 KB
- moon.yml 1 KB
- references/access-patterns.md 3 KB
- references/cache-behavior.md 3 KB
- references/data-as-transforms.md 3 KB
- references/existence-based-processing.md 3 KB
- references/handles-and-indices.md 3 KB
- references/measurement-and-profiling.md 3 KB
- references/nested-arrays.md 3 KB
- references/simd-friendly-layout.md 3 KB
- references/soa-aos-aosoa.md 3 KB
- references/statistics-recording.md 3 KB
- references/zero-as-default.md 2 KB
Wersje
W bazie nie ma innych skilli o nazwie data-oriented-design-guide.
Migawka bazy z 13.11.2026 · wersja paczki e2adf2905c94 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 100cpu-cache-opt
Optymalizuje użycie pamięci podręcznej CPU w C, C++ i Rust: diagnoza chybień, układ danych, liczniki perf stat, false sharing, prefetching i porównanie układów AoS oraz SoA. - 100cpu-cache-opt
Pomaga diagnozować chybienia cache przez perf, usuwać false sharing, wybierać układ AoS lub SoA i dodawać programowy prefetch. - 100array
Przegląda i przepisuje gorący kod według projektowania zorientowanego na dane: płaskie tablice, indeksy zamiast wskaźników i bitsety. - 95aos-update
Sprawdza, pobiera i weryfikuje aktualizację zainstalowanego AOS, z podglądem zmian przed uruchomieniem samej aktualizacji. - 100aos-project-init
Przeprowadza wywiad z folderem projektu i zakłada w nim strukturę AOS: plik konfiguracji, pamięć, mapę i plik AGENTS.md. - 100data-oriented-perf
Przyspiesza kod krytyczny: najpierw profiluje, potem zmienia układ danych na płaskie tablice i bitsety, a każdą zmianę mierzy przed i po.