AVX

Assess

Techniques

x86 vector instruction set extension for accelerating data-parallel workloads.

Why it's here

Placed in Assess: 3 article(s) of evidence from 2 source(s), led by open-source activity, with 2 in the last 30 days. Confidence 47%.

7Hacker News·9/25/2026framework_update
Go adds experimental portable SIMD APIs
Go 1.26 and 1.27 introduce experimental SIMD APIs, including architecture-specific support for amd64, arm64, and wasm. The new portable simd package aims to let developers write platform-agnostic vectorized code with near-assembly performance, while also providing emulation on systems without native SIMD support.
7The New Stack·9/17/2026research
Intel Compresses Ternary LLM Weights Below 1.58 Bits with BITCOS
Intel researchers introduced BITCOS, a storage format that compresses ternary model weights by separating zero positions from sign bits, without changing the model’s weights or outputs. In tests across multiple checkpoints and hardware platforms, it reduced storage to as low as 1.485 bits per weight and improved decoding throughput by up to 27%.
5Hacker News·7/15/2026open_source
Gemma 4 runs on a 13-year-old Xeon with CPU-only optimizations
A Hacker News post describes running Google's Gemma 4 26B mixture-of-experts model on a repurposed 13-year-old HP server with dual Ivy Bridge Xeons and no GPU. After adjusting the inference code to handle pre-AVX2 CPUs, the model reaches about 5.2 tokens per second on the basement machine.