{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T14:49:54Z","timestamp":1785336594405,"version":"3.55.0"},"reference-count":46,"publisher":"Association for Computing Machinery (ACM)","issue":"9","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Proc. VLDB Endow."],"published-print":{"date-parts":[[2021,5]]},"abstract":"<jats:p>Graph Neural Networks (GNNs) are proven to be powerful models to generate node embedding for downstream applications. However, due to the high computation complexity of GNN inference, it is hard to deploy GNNs for large-scale or real-time applications. In this paper, we propose to accelerate GNN inference by pruning the dimensions in each layer with negligible accuracy loss. Our pruning framework uses a novel LASSO regression formulation for GNNs to identify feature dimensions (channels) that have high influence on the output activation. We identify two inference scenarios and design pruning schemes based on their computation and memory usage for each. To further reduce the inference complexity, we effectively store and reuse hidden features of visited nodes, which significantly reduces the number of supporting nodes needed to compute the target embedding. We evaluate the proposed method with the node classification problem on five popular datasets and a real-time spam detection application. We demonstrate that the pruned GNN models greatly reduce computation and memory usage with little accuracy loss. For full inference, the proposed method achieves an average of 3.27X speedup with only 0.002 drop in F1-Micro on GPU. For batched inference, the proposed method achieves an average of 6.67X speedup with only 0.003 drop in F1-Micro on CPU. To the best of our knowledge, we are the first to accelerate large scale real-time GNN inference through channel pruning.<\/jats:p>","DOI":"10.14778\/3461535.3461547","type":"journal-article","created":{"date-parts":[[2021,10,22]],"date-time":"2021-10-22T22:22:49Z","timestamp":1634941369000},"page":"1597-1605","source":"Crossref","is-referenced-by-count":52,"title":["Accelerating large scale real-time GNN inference using channel pruning"],"prefix":"10.14778","volume":"14","author":[{"given":"Hongkuan","family":"Zhou","sequence":"first","affiliation":[{"name":"University of Southern California"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ajitesh","family":"Srivastava","sequence":"additional","affiliation":[{"name":"University of Southern California"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hanqing","family":"Zeng","sequence":"additional","affiliation":[{"name":"University of Southern California"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rajgopal","family":"Kannan","sequence":"additional","affiliation":[{"name":"US Army Research Lab"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Viktor","family":"Prasanna","sequence":"additional","affiliation":[{"name":"University of Southern California"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2021,10,22]]},"reference":[{"key":"e_1_2_1_1_1","volume-title":"Greg Ver Steeg, and Aram Galstyan","author":"Abu-El-Haija Sami","year":"2019"},{"key":"e_1_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3403296"},{"key":"e_1_2_1_3_1","volume-title":"Tie yan Liu, and Liwei Wang","author":"Cai Tianle","year":"2020"},{"key":"e_1_2_1_4_1","volume-title":"International Conference on Learning Representations (ICLR).","author":"Chen Jie","year":"2018"},{"key":"e_1_2_1_5_1","unstructured":"Jianfei Chen Jun Zhu and Le Song. 2018. Stochastic Training of Graph Convolutional Networks with Variance Reduction.. In ICML. 941--949.  Jianfei Chen Jun Zhu and Le Song. 2018. Stochastic Training of Graph Convolutional Networks with Variance Reduction.. In ICML. 941--949."},{"key":"e_1_2_1_6_1","volume-title":"Proceedings of the IEEE conference on computer vision and pattern recognition. 652--660","author":"Qi Charles R C.","year":"2017"},{"key":"e_1_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01240-3_25"},{"key":"e_1_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.556"},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1145\/3292500.3330673"},{"key":"e_1_2_1_10_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00011"},{"key":"e_1_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.3301922"},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3403302"},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.5555\/3294771.3294869"},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.155"},{"key":"e_1_2_1_15_1","unstructured":"Weihua Hu Matthias Fey Marinka Zitnik Yuxiao Dong Hongyu Ren Bowen Liu Michele Catasta and Jure Leskovec. 2020. Open Graph Benchmark: Datasets for Machine Learning on Graphs. arXiv:2005.00687 [cs.LG]  Weihua Hu Matthias Fey Marinka Zitnik Yuxiao Dong Hongyu Ren Bowen Liu Michele Catasta and Jure Leskovec. 2020. Open Graph Benchmark: Datasets for Machine Learning on Graphs. arXiv:2005.00687 [cs.LG]"},{"key":"e_1_2_1_16_1","volume-title":"Kipf and Max Welling","author":"Thomas","year":"2017"},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1145\/3357384.3357820"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1186\/s13640-019-0476-x"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.1145\/3269206.3272010"},{"key":"e_1_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2018.2858232"},{"key":"e_1_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.1145\/2783258.2783370"},{"key":"e_1_2_1_22_1","volume-title":"SIGN: Scalable Inception Graph Neural Networks. CoRR","author":"Rossi Emanuele","year":"2020"},{"key":"e_1_2_1_23_1","volume-title":"Few-Shot Learning with Graph Neural Networks. In International Conference on Learning Representations.","author":"Satorras Victor Garcia","year":"2018"},{"key":"e_1_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-93417-4_38"},{"key":"e_1_2_1_25_1","volume-title":"GraphR: Accelerating Graph Processing Using ReRAM. In 2018 IEEE International Symposium on High Performance Computer Architecture (HPCA). 531--543","author":"Song L."},{"key":"e_1_2_1_26_1","volume-title":"PCGCN: Partition-Centric Processing for Accelerating Graph Convolutional Network. In 2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS). 936--945","author":"Tian C."},{"key":"e_1_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.2517-6161.1996.tb02080.x"},{"key":"e_1_2_1_28_1","volume-title":"Graph Attention Networks. International Conference on Learning Representations","author":"Veli\u010dkovi\u0107 Petar","year":"2018"},{"key":"e_1_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM.2019.00070"},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1145\/3326362"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.5555\/3157096.3157329"},{"key":"e_1_2_1_32_1","volume-title":"Proceedings of the 36th International Conference on Machine Learning. PMLR, 6861--6871","author":"Wu Felix","year":"2019"},{"key":"e_1_2_1_33_1","volume-title":"International Conference on Learning Representations.","author":"Xu Keyulu","year":"2019"},{"key":"e_1_2_1_34_1","volume-title":"Proceedings of the 35th International Conference on Machine Learning, ICML 2018, Stockholmsm\u00e4ssan","volume":"80","author":"Xu Keyulu","year":"2018"},{"key":"e_1_2_1_35_1","volume-title":"Dynamically Pruned Message Passing Networks for Large-scale Knowledge Graph Reasoning. In International Conference on Learning Representations.","author":"Xu Xiaoran","year":"2020"},{"key":"e_1_2_1_36_1","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3403236"},{"key":"e_1_2_1_37_1","volume-title":"HyGCN: A GCN Accelerator with Hybrid Architecture. In 2020 IEEE International Symposium on High Performance Computer Architecture (HPCA). 15--29","author":"Yan M."},{"key":"e_1_2_1_38_1","volume-title":"Rethinking the Smaller-Norm-Less-Informative Assumption in Channel Pruning of Convolution Layers. In International Conference on Learning Representations.","author":"Ye Jianbo"},{"key":"e_1_2_1_39_1","doi-asserted-by":"publisher","DOI":"10.1145\/3219819.3219890"},{"key":"e_1_2_1_40_1","volume-title":"GraphACT. The 2020 ACM\/SIGDA International Symposium on Field-Programmable Gate Arrays (Feb","author":"Zeng Hanqing","year":"2020"},{"key":"e_1_2_1_41_1","volume-title":"Efficient and Scalable Graph Embedding. In 2019 IEEE International Parallel and Distributed Processing Symposium (IPDPS).","author":"Zeng Hanqing","year":"2019"},{"key":"e_1_2_1_42_1","volume-title":"GraphSAINT: Graph Sampling Based Inductive Learning Method. In International Conference on Learning Representations.","author":"Zeng Hanqing","year":"2020"},{"key":"e_1_2_1_43_1","volume-title":"Hardware Acceleration of Large Scale GCN Inference. In 2020 IEEE 31st International Conference on Application-specific Systems, Architectures and Processors (ASAP). 61--68","author":"Zhang B."},{"key":"e_1_2_1_44_1","doi-asserted-by":"publisher","DOI":"10.5555\/3367471.3367634"},{"key":"e_1_2_1_45_1","volume-title":"Graph-Bert: Only Attention is Needed for Learning Graph Representations. arXiv preprint arXiv:2001.05140","author":"Zhang Jiawei","year":"2020"},{"key":"e_1_2_1_46_1","volume-title":"Manning","author":"Zhang Yuhao","year":"2018"}],"container-title":["Proceedings of the VLDB Endowment"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.14778\/3461535.3461547","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T09:48:59Z","timestamp":1672220939000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.14778\/3461535.3461547"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,5]]},"references-count":46,"journal-issue":{"issue":"9","published-print":{"date-parts":[[2021,5]]}},"alternative-id":["10.14778\/3461535.3461547"],"URL":"https:\/\/doi.org\/10.14778\/3461535.3461547","relation":{},"ISSN":["2150-8097"],"issn-type":[{"value":"2150-8097","type":"print"}],"subject":[],"published":{"date-parts":[[2021,5]]}}}