Dual RTX 3060 Faces MoE Loading Issues
esw123 · reddit · 2026-07-12
The author previously ran Qwen 3.6 27B, 35B, and 122B on a single RTX 3060 with CPU offload.
After adding a second 3060, the situation worsened: dense models load normally, 35B loads only once, 122B fails entirely. They want to know which settings to check or if there's a fix for dual-GPU loading issues.
More from Infra
- NVIDIA publishes Vera CPU architecture details before AMD’s AI event — ryanshrout · 2026-07-22
- oMLX 0.5.2 adds Mac menu-bar stats, low-bit decode kernels, and faster downloads — awnihannun · 2026-07-22
- Strangeworks launches Aura to turn enterprise ops into production optimization systems — whurley · 2026-07-22
- Graph workload 854.graph500 enters SPEC CPU 2026 as a new CPU benchmark — Prof_DavidBader · 2026-07-22
- HilbertRaum open-sources a fully local AI chat and document analysis app for private use — Vladowski · 2026-07-22
- Hybrid and local inference are emerging as a response to AI energy and token costs — dmitry140 · 2026-07-22