Prefill vs decode in LLM inference Fanout ProContinue readingThis source-rich field note is available with Fanout Pro.Get Fanout ProRelated articlesWhat is chunked prefill?Continuous batching for LLM inferenceAI inference engineering, explained with numbersContinue learningInference Memory and KV CacheInference engineering