2 papers
cs.DC2026
GPUOS: A GPU Operating System Primitive for Transparent Operation Fusion
Yiwei Yang, Xiangyu Gao, Yuan Zhou +3
Modern deep learning workloads often consist of many small tensor operations, especially in inference, attention, and micro-batched training. In these settings, kernel launch overh…
cs.LG2025
Verifying Computational Graphs in Production-Grade Distributed Machine Learning Frameworks
Kahfi S. Zulkifli, Wenbo Qian, Shaowei Zhu +3
Modern machine learning frameworks support very large models by incorporating parallelism and optimization techniques. Yet, these very techniques add new layers of complexity, intr…