FIELD GUIDES
Guides
Decision-focused explainers built from the lab's first-party evidence. No textbook rewrites — each guide ties to real benchmark numbers.
The local-AI GPU buyer's guide
12 minHow to choose between NVIDIA, AMD, and Intel Arc for local LLM workloads — by workload, not by brand loyalty.
Quantization, honestly: Q4 vs Q6 vs Q8 vs FP8
9 minWhat you actually lose when you shrink a model. Real quality and speed deltas, not the textbook version.
Vulkan vs SYCL vs CUDA: the backend decision
11 minWhy Intel Arc Battlemage ships on Vulkan, what CUDA still wins, and where AMD lands.
Will it fit? VRAM math for local LLMs
7 minThe arithmetic behind Model Fit, with worked examples from 7B to 70B.