🖥️ Modern CPU - matmat · Scour

KTransformers Adds AVX2 MoE Support For Viable Performance On CPUs Without AMX/AVX-512 ⚡SIMD Optimization

phoronix.com·15h·

AMD's Ryzen 9 9950X3D2 Dual Edition crams 208MB of cache into a single chip 🖥️Hardware Architecture

arstechnica.com·6d·Hacker News·

IsoQuant: Hardware-Aligned SO(4) Isoclinic Rotations for LLM KV Cache Compression 🖥️Hardware Architecture

arxiv.org·2d·

'Performance without compromise': AMD debuts first dual 3D V-Cache Ryzen CPU in potential showdown against Threadripper and EPYC siblings 🚀RISC-V Vector

techradar.com

·2d·

We test Intel's secret sauce for Windows 11 performance in its latest Core Ultra processors 🖥️Modern Terminals

neowin.net·4d·

Geekbench investigates up to 30% jump with Intel's iBOT — performance gain attributed to newly-vectorized instructions 🧮Compute Optimization

tomshardware.com

·2d·

Stack vs malloc: real-world benchmark shows 2–6x difference 🧠Memory Allocators

blog.stackademic.com

·1d·DEV·

Peterc3-dev/rag-race-router: R.A.G-Race-Router [Adaptive Tri-Processor Inference Runtime] — Self-optimizing CPU+iGPU+NPU inference for AMD Ryzen AI 300 series 🖥️Hardware Architecture

github.com·3d·Hacker News·

Intel Binary Optimization Tool Changes Code Execution with Heavy Vectorization 🚀Compiler Optimizations

techpowerup.com·2d·

Metal Quantized Attention: pulling M5 Max ahead with Int8 matrix multiplication 📊RISC-V Vectors

releases.drawthings.ai·1d·Hacker News·

DPU Vivado + Petalinux flow 2025.2 ⚡Homebrew CPUs

hackster.io·6d·

WheatForce: Learning From CPU Architecture Mistakes 🖥️Hardware Architecture

hackaday.com·1d·

A n00b PM’s guide to vibe coding kernels from scratch 🔩Systems Programming

ddmckinnon.com·4d·Hacker News·

TurboQuant: Building a Sub-Byte KV Cache Quantizer from Paper to Production ⚡Homebrew CPUs

demo.aitherium.com·6d·Hacker News·

TurboQuant tackles the hidden memory problem that's been limiting your local LLMs 💻Local LLMs

xda-developers.com·3d·

KV Cache Quantization for Self-Forcing Video Generation: A 33-Method Empirical Study 🖥️Hardware Architecture

arxiv.org·2d·

bloderxd/kotlin-cpu: 16-bit CPU emulator in Kotlin 🎯Emulator Accuracy

github.com·1d·Hacker News·

Intel Announces The "Optimization Zone" 🖥️Hardware Architecture

phoronix.com·3d·

KVSculpt: KV Cache Compression as Distillation ⚡Cache Theory

arxiv.org·2d·

Essenceia/uselessly_fast_bfloat16_multiplier: Pushing the bf16 multiplication clock frequency to the max on the nominal corner on IHP 130nm 5L node. ⚡Homebrew CPUs

github.com·4d·Hacker News·

Loading more...