RTX Spark vs M5 Max: Local Qwen 3.8 27B Inference Benchmark

A user benchmarked Qwen 3.8 27B locally on NVIDIA's RTX Spark N1X versus Apple's M5 Max, finding that while the M5 Max has higher memory bandwidth, its compute falls well short of NVIDIA's offering.

2026-10-19 ~ 2026-10-19 · 2 related posts

• RTX Spark vs M5 Max: local Qwen 3.8 27B benchmark shows NVIDIA wins on compute — draginol · 2026-10-19
• Follow-up: M5 Max has higher bandwidth but compute falls well short of RTX Spark — draginol · 2026-10-19