Glossary

Best Computer for Local LLM: glossary

Last updated 13 November 2026

Answer in brief

Eight terms that decide which machine runs which model.

Olares One → Computer for local LLMs → Ollama · vLLM · LM Studio → Local models without the cloud

Memory on the graphics card. A model runs fastest when all of its weights fit in VRAM; whatever doesn't fit spills to sy

Memory shared by the CPU and GPU in one pool, as on Apple's M-series chips, NVIDIA GB10 (DGX Spark) and AMD Ryzen AI Max

Storing a model's weights at lower precision, for example 4-bit instead of 16-bit, so it needs about a quarter of the me

How fast a model writes its answer. A token is roughly three quarters of an English word; reading speed is around 5 to 1

How much text a model can consider at once: the prompt, documents and the conversation so far. Longer contexts need more

A model built from many smaller expert networks, of which only a few run for each token. Qwen3-30B-A3B has about 30B par

NVIDIA's platform for running software on its GPUs. Most AI tools, including vLLM and most image and video generators, a

Letting a model answer from your own documents: the system searches the documents, passes the relevant passages to the m

Sources for this entry

Best Computer for Local LLM. “Best Computer for Local LLM: glossary.” Content reviewed 13 November 2026.