Description

A custom portable inference setup using dual M3 Ultra Mac Studios linked via Thunderbolt 5 to achieve high-speed distributed LLM execution. The system achieves 203 tokens per second running deepseek-v4.1-flash with a time-to-first-token of 0.18s, demonstrating seamless data transfer across the high-bandwidth link.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.