Real-Time On-Device Diffusion: Practical Acceleration via Fused Low-Bit Kernels
A systems paper on accelerating diffusion inference with fused low-bit kernels and cache-update fusion.
On-Device Diffusion Kernels
Kernel-level work on MoDiff: fused low-bit operators and cache-update fusion for measured, not only counted, diffusion speedups.
•
1 min read
Read more