Tag

hardware

All blog posts with this tag.

6 Oct 2026
Local MoE LLMs without a server GPU: RAM, VRAM, and what you need in 2026
Why Mixture of Experts needs huge memory even with few active parameters, how offload and quantization change the home build, and why 64 GB RAM + 16 GB VRAM is a serious local AI start.
AI hardware llm local ai Tutorial
1 Oct 2026
Local LLM on an NPU mini-PC: what Ryzen AI 9 365 actually delivered
A Habr field report: Qwen3.6–35B on FastFlowLM, RAM traps, one-request-at-a-time NPU limits, and payback versus cloud APIs.
AI News hardware
29 Aug 2026
CMP 90HX for LLM: why a mining GPU slows down and what “six-seven” fixes
MoE tolerates CMP 90HX; dense models don't — compute paths are capped. Driver mod via V67 nearly doubles throughput but doesn't erase mining-card limits.
AI News gpu hardware llm