8 papers
SparseDitto: An Agentic Sparse Compilation Framework through Architecture-Aware Synthesis on GPUs
Shiyang Li, Guangyan Sun, Jinwei Tang +3
Sparse matrix computation performance on GPU depends on how representation and execution schedule match the input structure and target hardware. No single implementation consistent…
CUDABeaver: Benchmarking LLM-Based Automated CUDA Debugging
Shiyang Li, Haoyang Chen, Mattia Fazzini +1
Debugging CUDA programs has long been challenging because failures often arise from subtle interactions among hardware behavior, compiler decisions, memory hierarchy, and asynchron…
CUDAHercules: Benchmarking Hardware-Aware Expert-level CUDA Optimization for LLMs
Shiyang Li, Zijian Zhang, Guangyan Sun +5
Large language models show promise for automated CUDA programming, however even the strongest coding models (e.g., Claude-Opus-4.6) may still fall short of expert-level, architectu…
FPTC: A Fast Parallel Transform-based Codec for Efficient Asymmetric Signal Compression
Ben Mechels, Ryan Billmeyer, Alexander Chen +2
Modern high-performance computing and Internet-of-Things deployments increasingly generate large volumes of signal data that must be compressed efficiently on resource-constrained…
GSR-GNN: Training Acceleration and Memory-Saving Framework of Deep GNNs on Circuit Graph
Yuebo Luo, Shiyang Li, Yifei Feng +3
Graph Neural Networks (GNNs) show strong promise for circuit analysis, but scaling to modern large-scale circuit graphs is limited by GPU memory and training cost, especially for d…
StitchCUDA: An Automated Multi-Agents End-to-End GPU Programing Framework with Rubric-based Agentic Reinforcement Learning
Shiyang Li, Zijian Zhang, Winson Chen +3
Modern machine learning (ML) workloads increasingly rely on GPUs, yet achieving high end-to-end performance remains challenging due to dependencies on both GPU kernel efficiency an…
CudaForge: An Agent Framework with Hardware Feedback for CUDA Kernel Optimization
Zijian Zhang, Rong Wang, Shiyang Li +3
Developing efficient CUDA kernels is increasingly critical for AI applications such as large-scale LLM training. However, manual kernel design is both costly and time-consuming, mo…
DR-CircuitGNN: Training Acceleration of Heterogeneous Circuit Graph Neural Network on GPUs
Yuebo Luo, Shiyang Li, Junran Tao +6
The increasing scale and complexity of integrated circuit design have led to increased challenges in Electronic Design Automation (EDA). Graph Neural Networks (GNNs) have emerged a…