Local LLMs on Refurb M4 Max vs new M5 Max
A data scientist is seeking advice on whether to purchase a refurbished MacBook Pro with an M4 Max chip or a new MacBook Pro with an M5 Max chip for running local large language models. The M5 Max offers a slight increase in CPU cores and memory bandwidth, potentially improving performance for tasks like data derivation and extraction with models such as Gemma 4 31B and Qwen 3.6-27B. The decision hinges on whether the $1,120 price difference justifies the incremental performance gains of the newer chip. AI
IMPACT Users considering local LLM deployment will weigh hardware specs like CPU, GPU, and memory bandwidth for performance.