🔀 SIMD Programming - hello · Scour

Automatic Vectorization ⚡SIMD Optimization

en.wikipedia.org·2d·Hacker News

CUTEv2: Unified and Configurable Matrix Extension for Diverse CPU Architectures with Minimal Design Overhead ⚡Hardware Acceleration

arxiv.org·18h

Data Oriented Design by example (2017) ⚡Cache Optimization

nikitablack.github.io·4d·Hacker News

Rust and AI: Building the Next Generation of High-Performance Machine Learning Systems 🍱Nom

levelup.gitconnected.com

·1d

naresh-cn2/Axiom-Turbo-IO: High-performance C-engine for parallel data processing. Parses 10M rows in 0.08s using multithreaded mmap. 🌀Naiad

github.com·3d·DEV

Boost Your Spark Jobs: How Photon Accelerates Apache Spark Performance 💡Photon

dzone.com·1d

AEG: A Baremetal Framework for AI Acceleration via Direct Hardware Access in Heterogeneous Accelerators ⚡Hardware Acceleration

arxiv.org·18h

Interferences within a certifiable design methodology for high-performance multi-core platforms 🧵OpenMP

arxiv.org·18h

Leveraging Mathematical Reasoning of LLMs for Efficient GPU Thread Mapping 🎮SIMT Execution

arxiv.org·18h

Sparsity-Aware Roofline Models for Sparse Matrix-Matrix Multiplication 🔢Sparse Matrices

arxiv.org·5d

SPEED-Bench: A Unified and Diverse Benchmark for Speculative Decoding 📦Folly

arxiv.org·18h

Benchmarking Compound AI Applications for Hardware-Software Co-Design 🚀Performance

arxiv.org·18h

A Full-Stack Performance Evaluation Infrastructure for 3D-DRAM-based LLM Accelerators 🌊Memory Bandwidth

arxiv.org·4d

Tessera: Unlocking Heterogeneous GPUs through Kernel-Granularity Disaggregation 🚀Milvus

arxiv.org·18h

WaveTune: Wave-aware Bilinear Modeling for Efficient GPU Kernel Auto-tuning 💡Photon

arxiv.org·18h

Wattlytics: A Web Platform for Co-Optimizing Performance, Energy, and TCO in HPC Clusters 🏗️System Design

arxiv.org·4d

Technology solutions targeting the performance of gen-AI inference in resource constrained platforms 🧩mimalloc

arxiv.org·18h

Making Room for AI: Multi-GPU Molecular Dynamics with Deep Potentials in GROMACS 🚀Milvus

arxiv.org·5d

Analyzing Persistent Alltoallv RMA Implementations for High-Performance MPI Communication ⚡RDMA

arxiv.org·6d

SepSeq: A Training-Free Framework for Long Numerical Sequence Processing in LLMs 🌀Naiad

arxiv.org·4d