GPUs are critical for compute-intensive applications, yet emerging workloads such as recommender systems, graph analytics, and data analytics often exceed GPU memory capacity. Existing solutions allow GPUs to use CPU DRAM or SSDs as external memory, and the GPU-centric approach enables GPU threads to directly issue NVMe requests, further avoiding CPU intervention. However, current GPU-centric approaches adopt synchronous I/O, forcing threads to stall during long communication delays. We propose AGILE, a lightweight asynchronous GPU-centric I/O library that eliminates deadlock risks and integrates a flexible HBM-based software cache. AGILE overlaps computation and I/O, improving performance by up to 1.88× across workloads with diverse computation-to-communication ratios. Compared to BaM on DLRM, AGILE achieves up to 1.75× speedup through efficient design and overlapping; on graph applications, AGILE reduces software cache overhead by up to 3.12× and NVMe I/O overhead by up to 2.85×; AGILE also lowers per-thread register usage by up to 1.32×.
@article{arxiv.2504.19365,
title = {AGILE: Lightweight and Efficient Asynchronous GPU-SSD Integration},
author = {Zhuoping Yang and Jinming Zhuang and Xingzhen Chen and Alex K. Jones and Peipei Zhou},
journal= {arXiv preprint arXiv:2504.19365},
year = {2025}
}