Glossary
Short, plain-language explainers for the terms you’ll meet across the site.
- Airflow explained
- Memory bandwidth explained
- Quantization explained
- RAM explained
- Unified memory explained
- What are local LLMs?
- What are model parameters?
- What are model weights?
- What are tokens per second?
- What does 70B mean?
- What does TOPS mean?
- What is a 70B model?
- What is a chassis?
- What is a context window?
- What is a CPU?
- What is a GPU?
- What is a local LLM?
- What is a mini PC?
- What is a model's memory footprint?
- What is a prompt?
- What is a token?
- What is an eGPU?
- What is an iGPU?
- What is an LLM?
- What is an NPU?
- What is an SSD?
- What is DDR5?
- What is dual-channel memory?
- What is FP16?
- What is GB/s?
- What is inference?
- What is llama.cpp?
- What is memory headroom?
- What is OCuLink?
- What is PCIe?
- What is prompt processing?
- What is SO-DIMM memory?
- What is Strix Halo?
- What is tok/s?
- What is USB-C?