Abstract:Diffusion models have become essential for high-fidelity 3D MRI synthesis, yet their deployment remains constrained by substantial GPU resource demands arising from hundreds of U-Net evaluations per sample and a highly heterogeneous kernel behavior. This paper performs a comprehensive performance analysis of the state-of-the-art medical diffusion model, Med-DDPM, across three generations of NVIDIA architectures to study kernel-level runtime breakdowns, instruction-mix characteristics, memory system utilization, warp-level activities, and profiler priority-score estimates. We show that training is overwhelmingly dominated by cuDNN convolution and implicit-GEMM kernels, with inefficiencies arising from memory-access patterns, tensor-layout conversions, and limited Tensor Core utilization. Guided by these insights, we evaluate two architecture-aware optimizations TF32 Tensor Core activation and a 3D channels-last layout and demonstrate that they reduce SM cycles by up to 100x, cut dynamic instructions by 100x, raise Tensor Core utilization from 1.45 to 9.98x, and increase IPC by 7% on A100, all without degrading synthesis quality.
From: Jeeho Ryoo [view email]
[v1]
Thu, 11 Jun 2026 02:12:10 UTC (438 KB)