#simd — HitReader

performance engineering Sep 16, 2026 6 min read

How Vectorized Quicksort Fills Every CPU Lane

Vectorized Quicksort keeps the familiar pivot-and-partition strategy but feeds many values through CPU vector lanes at once. This article explains SIMD partitioning, compress-store operations, sorting networks, Highway portability, and the trade-offs behind the speedup.

by ahsan

#c++ #cpu performance #highway #quicksort #simd

machine learning Aug 18, 2026 6 min read

Turbovec: Google’s TurboQuant Vector Search in Rust

Turbovec is a Rust vector index built on Google’s TurboQuant. It stores high-dimensional embeddings in 2–4 bits per coordinate, avoids a separate training phase, and uses SIMD-friendly layouts for fast similarity search.

by ahsan

#machine learning #quantization #rust #simd #vector search