Systematic benchmarking of diffusion model optimizations on Apple M3 Ultra produces 22.7 FPS real-time img2img at 512x512 and demonstrates that CUDA-derived techniques do not transfer directly to Apple Silicon.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Systematic Optimization of Real-Time Diffusion Model Inference on Apple M3 Ultra
Systematic benchmarking of diffusion model optimizations on Apple M3 Ultra produces 22.7 FPS real-time img2img at 512x512 and demonstrates that CUDA-derived techniques do not transfer directly to Apple Silicon.