SeedVR2 Video Upscaler v1.5.5 Brings TensorRT to ComfyUI, with Big Decoder Gains but Little Encoder Speedup
Zestyclose_Bake3680 · reddit · 2026-09-28
- Released ComfyUI-SeedVR2-VideoUpscaler-with-TensorRT v1.5.5: the SeedVR2 VAE Encoder now runs via TensorRT inside ComfyUI. Porting was hard because ComfyUI's custom VRAM management clashes with tooling built for standalone use; the encoder previously produced blackouts and noise.
- Includes an example workflow JSON and a TRT engine build for Blackwell GPUs hosted on Hugging Face.
- Practical takeaway: TensorRT speedups are pronounced on the decoder side, but the encoder shows little gain over fp16 — in some cases fp16 is even faster.
More from Infra
- Anthropic's Contracted Value Exceeds $180B, Per NVIDIA — zephyr_z9 · 2026-09-28
- David Friedberg: Open-Source Models Make Data Centers Unnecessary for 90% of AI Use — MarvinTBaumann · 2026-09-28
- Old-School Infra Engineers on PB-Scale AI Logs: Hadoop Did This on a Tuesday — basedjensen · 2026-09-28
- Power sharing fits 192 GB300 GPUs in 140-GPU power budget, boosting tokens/s by 49% — julsimon · 2026-09-28
- Node.js async zstdCompress() was 23x slower due to missing pledged input size — AI Snake Oil · 2026-09-28
- Category theory meets deep learning: PyNCD diagrams derive hardware-aware FlashAttention — GioeleZardini · 2026-09-28