SparseDitto: An Agentic Sparse Compilation Framework through Architecture-Aware Synthesis on GPUs

תקציר מקורי באנגליתarXiv:2608.05033v3 Announce Type: replace-cross Abstract: Sparse matrix computation performance on GPU depends on how representation and execution schedule match the input structure and target hardware. No single implementation consistently dominates across sparsity patterns, operators, and hardwares. Existing sparse compilers and specialized systems cannot cover all of them simultaneously. We present SparseDitto, an agentic sparse compilation framework for sparse matrix computation on GPUs. It jointly synthesizes representation, execution schedule, and hardware mapping in a unified compilation plan. Structural analysis and a learned template-ranking prior guide architecture-aware synthesis. LLM-guided lowering realizes each plan as CUDA code, while target-GPU profiling drives plan refinem