RTX 5090 ComfyUI: Massive Performance Leap Migrating from Windows to Linux
No_Statement_7481 · reddit · 2026-07-20
A user running an RTX 5090 (95GB RAM) shared their detailed experience migrating ComfyUI workflows from Windows to Linux (Kubuntu), noting that recent ComfyUI updates caused severe VRAM management regressions on Windows.
Windows Pain Points:
- Workflows frequently encountered OOM (Out of Memory) errors or massive slowdowns.
- The system seemed to throttle the GPU's high VRAM advantage, leading to slow swapping and poor generation times.
- Dependency management turned into "dependency hell," requiring environment fixes after every update.
Linux Migration Improvements:
- Speed Surge: Single-step generation time for LTX 2.3 workflows plummeted from 15-20 seconds to 2 seconds.
- Stability Restored: The OOM issues with Wan 2.2 completely vanished.
- Unleashed Hardware Potential: Processing speeds for nodes like RifeVFI and AnimateDiff improved dramatically, finally hitting the 5090's target performance.
The author concludes that for users heavily reliant on video generation and ComfyUI, switching to Linux is absolutely worth it if they feel their high-end GPU is underperforming on Windows.
More from Infra
- CoreWeave 5-year CDS spikes to ~1,100bps as market fears AI bubble starts as credit event — kimmonismus · 2026-10-26
- NVIDIA Plans ~$2,000 Per GPU Per Year Software Fee for Running AI Agents — LinusEkenstam · 2026-10-26
- 12 DevOps concepts to learn in order, from Linux to SRE — _jaydeepkarale · 2026-10-26
- Killed our Vector DB for SQLite BM25 and wrote a 3.6ns Rust sentinel: lessons from multi-agent loops — OpenComplaint524 · 2026-10-26
- Gemma 4 26B-A4B runs offline on a 16 GB Mac: 33/40 on TypeScript tasks at 31 tok/s — Mission-Sir-1614 · 2026-10-26
- How thousands of GPUs train one AI: a breakdown of 5D parallelism — jbhuang0604 · 2026-10-26