4 dépôts
Strategies for structuring computations to minimize data movement between memory hierarchies.
Distinct from Data Layout Optimizers: Distinct from query layout: focuses on cache-aware memory access for numerical algorithms.
Explore 4 awesome GitHub repositories matching data & databases · Data Locality Optimizers. Refine with filters or upvote what's useful.
This project is a comprehensive library for numerical linear algebra and scientific computing, designed to provide optimized routines for matrix decomposition, statistical modeling, and high-performance data analysis. It serves as both a toolkit for solving complex linear systems and an educational resource for understanding the fundamental algorithms behind matrix factorizations and numerical solvers. The library distinguishes itself through a focus on randomized numerical linear algebra, utilizing probabilistic algorithms and approximate methods to perform dimensionality reduction and matri
Structures computations to minimize data movement between memory hierarchies for quicker retrieval.
Odin is a compiled, statically typed systems programming language designed for high-performance software development. It focuses on pragmatic low-level memory control, providing a toolset for manual memory management and precise control over hardware utilization. The language is distinguished by its flexible memory model, which includes custom allocators and precise data layout capabilities to optimize resource usage. It features a comprehensive foreign function interface for importing assembly files and linking with external libraries using configurable calling conventions. The type system
Supports organizing records as arrays of structures or structures of arrays to maximize hardware acceleration.
Thorium is a web browser built from the Chromium project, designed for high performance and expanded compatibility. It utilizes aggressive compiler optimizations and CPU-specific instruction sets, such as AVX2 and SIMD, to increase page rendering and JavaScript execution speeds. The project distinguishes itself by providing custom builds that enable modern web browsing on legacy versions of Windows and Linux. It further diverges from standard browser implementations by integrating Widevine DRM and native support for high-efficiency media formats, including HEVC and JPEG XL. Broad capabilitie
Implements strategies to reduce execution overhead by enhancing data locality within software loops.
Mooncake est une plateforme de service de modèles de langage (LLM) désagrégée et un magasin clé-valeur distribué conçu pour une infrastructure d'inférence haute performance. Il fonctionne comme un orchestrateur de mémoire GPU et un système de gestion de cache KV qui mutualise et transfère les caches clé-valeur à travers les clusters pour accélérer l'inférence. Le système se distingue en séparant les phases de pré-remplissage (prefill) et de décodage de l'inférence dans des clusters matériels distincts pour optimiser l'utilisation des ressources. Il utilise un cache distribué RDMA haute performance avec des transferts zéro-copie pour déplacer les données entre les nœuds de calcul, contournant le CPU pour réduire la latence et la surcharge. La plateforme couvre de vastes domaines de capacités, notamment la mutualisation de mémoire distribuée, le routage de mémoire d'accélérateur via CXL et le déchargement de stockage multi-niveaux vers des SSD. Il gère l'état du cluster via des services de coordination de métadonnées et implémente la gouvernance des ressources via une protection d'objets basée sur des baux et une éviction de cache basée sur des seuils. Le logiciel est packagé pour un déploiement conteneurisé avec prise en charge du réseau hôte et du mappage de périphériques matériels.
Assigns preferred storage segments for object allocation to minimize network overhead and increase speed.