Nunchaku Lite cuts peak VRAM by 50% and speeds image generation 30% in Diffusers

RisingSayak · x · 2026-07-24

Nunchaku Lite brings big image models to more GPUs

• Nunchaku Lite now works natively in Diffusers for SVDQuant 4-bit (W4A4) diffusion inference.
• The team says it cuts peak VRAM by up to 50% and makes generation about 30% faster without hurting image quality.
• The screenshot shows an example dropping from 31.1 GB to 16.0 GB peak memory, with end-to-end latency improving from 3.00s to 1.68s.
• The practical effect: larger image models become usable on many more GPUs.

Original post →

More from Infra

• RTK Terminal Compression Cuts Tokens but Leaves Your AI Coding Bill Unchanged — Bartaseth · 2026-09-11
• SF Compute founder: buying compute is 'an absolutely awful experience' right now — IgorCarron · 2026-09-11
• SmolVM open-sources persistent computer infrastructure for agents that outlive chat sessions — aniketmaurya · 2026-09-11
• PyTorch Day Korea 2026 launches first offline conf, CFP closes Sept 13 — PyTorch · 2026-09-11
• Local LLM server dilemma: 4x CMP-170HX (price up 53% in 20 days) vs Mac Studio M5 Ultra — rumboll · 2026-09-11
• llama.cpp lands Flash Attention tuning for RDNA4, big prefill gains on AMD — pmttyji · 2026-09-11

Infra channel →