ComfyUI Dual-GPU Pitfall: Second GPU Triggers OOM Errors
tricck3zz · reddit · 2026-08-13
A developer added a second GPU (RTX 3060 12GB) to their setup, which originally ran MiniMax workflows smoothly on an RTX 5070 Ti and 32GB RAM.
Without changing any node settings, the workflow now consistently throws an Out of Memory (OOM) error after running the text encoder. The post seeks answers for this abnormal VRAM allocation behavior in multi-GPU ComfyUI environments.
More from Infra
- Microsoft's Homegrown AI Chip Effort Shows Signs of Life After Slow Start — pstAsiatech · 2026-08-13
- Why Nvidia Is Trying To Develop the World's Best Open-Source AI Models — pstAsiatech · 2026-08-13
- Saving $40K in Monthly API Costs via Context Caching — NathanWilbanks_ · 2026-08-13
- LTX-Video Local Test: 30-Second Generation Takes 7 Minutes on RTX 5060 — Character_Title_876 · 2026-08-13
- CoreWeave's Profit Surprise Boosts AI Compute Confidence; Quantum Market Heads for Duopoly — TiernanRayTech · 2026-08-13
- LLM API Hosting Margins Squeezed Dry: Kimi K3 Pricing Hits Cost Floor — sergeykarayev · 2026-08-13