{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,28]],"date-time":"2026-04-28T04:13:02Z","timestamp":1777349582782,"version":"3.51.4"},"publisher-location":"New York, NY, USA","reference-count":33,"publisher":"ACM","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2025,9,8]]},"DOI":"10.1145\/3718958.3750539","type":"proceedings-article","created":{"date-parts":[[2025,8,27]],"date-time":"2025-08-27T16:54:11Z","timestamp":1756313651000},"page":"453-466","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Alibaba Stellar: A New Generation RDMA Network for Cloud AI"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-9902-585X","authenticated-orcid":false,"given":"Jie","family":"Lu","sequence":"first","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3548-2030","authenticated-orcid":false,"given":"Jiaqi","family":"Gao","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, USA"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2679-4983","authenticated-orcid":false,"given":"Fei","family":"Feng","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-8488-8556","authenticated-orcid":false,"given":"Zhiqiang","family":"He","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6615-0744","authenticated-orcid":false,"given":"Menglei","family":"Zheng","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-5266-8975","authenticated-orcid":false,"given":"Kun","family":"Liu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-9775-8591","authenticated-orcid":false,"given":"Jun","family":"He","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-4449-2436","authenticated-orcid":false,"given":"Binbin","family":"Liao","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-8856-2471","authenticated-orcid":false,"given":"Suwei","family":"Xu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-7031-9867","authenticated-orcid":false,"given":"Ke","family":"Sun","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-6682-1414","authenticated-orcid":false,"given":"Yongjia","family":"Mo","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-9383-5566","authenticated-orcid":false,"given":"Qinghua","family":"Peng","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-8299-6902","authenticated-orcid":false,"given":"Jilie","family":"Luo","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-0007-3828","authenticated-orcid":false,"given":"Qingxu","family":"Li","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-5353-3473","authenticated-orcid":false,"given":"Gang","family":"Lu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-6571-0539","authenticated-orcid":false,"given":"Zishu","family":"Wang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0939-8943","authenticated-orcid":false,"given":"Jianbo","family":"Dong","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3959-5086","authenticated-orcid":false,"given":"Kunling","family":"He","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4686-6120","authenticated-orcid":false,"given":"Sheng","family":"Cheng","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-5468-7366","authenticated-orcid":false,"given":"Jiamin","family":"Cao","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-4029-7421","authenticated-orcid":false,"given":"Hairong","family":"Jiao","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-6740-227X","authenticated-orcid":false,"given":"Pengcheng","family":"Zhang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-7057-0194","authenticated-orcid":false,"given":"Shu","family":"Ma","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-5145-6097","authenticated-orcid":false,"given":"Lingjun","family":"Zhu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0057-6678","authenticated-orcid":false,"given":"Chao","family":"Shi","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-6284-5231","authenticated-orcid":false,"given":"Yangming","family":"Zhang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-8714-6949","authenticated-orcid":false,"given":"Yiquan","family":"Chen","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2713-9075","authenticated-orcid":false,"given":"Wei","family":"Wang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-5404-797X","authenticated-orcid":false,"given":"Shuhong","family":"Zhu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-2725-9482","authenticated-orcid":false,"given":"Xingru","family":"Li","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2509-044X","authenticated-orcid":false,"given":"Qiang","family":"Wang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-0506-6634","authenticated-orcid":false,"given":"Jiang","family":"Liu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5676-102X","authenticated-orcid":false,"given":"Chao","family":"Wang","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3003-0150","authenticated-orcid":false,"given":"Wei","family":"Lin","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4352-7497","authenticated-orcid":false,"given":"Ennan","family":"Zhai","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7417-5469","authenticated-orcid":false,"given":"Jiesheng","family":"Wu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5792-322X","authenticated-orcid":false,"given":"Qiang","family":"Liu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-1213-0554","authenticated-orcid":false,"given":"Binzhang","family":"Fu","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-7272-8143","authenticated-orcid":false,"given":"Dennis","family":"Cai","sequence":"additional","affiliation":[{"name":"Alibaba Cloud, Sunnyvale, USA"}]}],"member":"320","published-online":{"date-parts":[[2025,8,27]]},"reference":[{"key":"e_1_3_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.1109\/49.103550"},{"key":"e_1_3_2_1_2_1","unstructured":"2024. NVIDIA GPUDirect RDMA. http:\/\/docs.nvidia.com\/cuda\/gpudirect-rdma\/. (Aug 2024)."},{"key":"e_1_3_2_1_3_1","volume-title":"PCI Express Specification. https:\/\/pcisig.com\/specifications. (2024)","year":"2024","unstructured":"2024. PCI Express Specification. https:\/\/pcisig.com\/specifications. (2024). Accessed: Sep 2024."},{"key":"e_1_3_2_1_4_1","volume-title":"Remote Direct Memory Accesss. (2024). https:\/\/en.wikipedia.org\/wiki\/Remote_direct_memory_access Accessed","year":"2025","unstructured":"2024. Remote Direct Memory Accesss. (2024). https:\/\/en.wikipedia.org\/wiki\/Remote_direct_memory_access Accessed: Jan 2025."},{"key":"e_1_3_2_1_5_1","unstructured":"2025. Nvidia Multi-Instance GPU. https:\/\/www.nvidia.com\/en-us\/technologies\/multi-instance-gpu\/"},{"key":"e_1_3_2_1_6_1","unstructured":"2025. Nvidia vGPU. https:\/\/docs.nvidia.com\/vgpu\/sizing\/virtual-workstation\/latest\/right-gpu.html"},{"key":"e_1_3_2_1_7_1","volume-title":"paravirtualization. https:\/\/www.techtarget.com\/searchitoperations\/definition\/paravirtualization. (Jan","year":"2025","unstructured":"2025. paravirtualization. https:\/\/www.techtarget.com\/searchitoperations\/definition\/paravirtualization. (Jan 2025)."},{"key":"e_1_3_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2017.12.007"},{"key":"e_1_3_2_1_9_1","volume-title":"2011 USENIX Annual Technical Conference (USENIX ATC 11)","author":"Amit Nadav","year":"2011","unstructured":"Nadav Amit, Muli Ben-Yehuda, IBM Research, Dan Tsafrir, and Assaf Schuster. 2011. vIOMMU: Efficient IOMMU Emulation. In 2011 USENIX Annual Technical Conference (USENIX ATC 11). USENIX Association, Portland, OR. https:\/\/www.usenix.org\/conference\/usenixatc11\/viommu-efficient-iommu-emulation"},{"key":"e_1_3_2_1_10_1","volume-title":"Rong Pan, Yanfang Le, Costin Raiciu, Mark Handley, Timo Schneider, Nils Blach, Ahmad Ghalayini, Daniel Alves, Michael Papamichael, Adrian Caulfield, and Torsten Hoefler.","author":"Bonato Tommaso","year":"2024","unstructured":"Tommaso Bonato, Abdul Kabbani, Daniele De Sensi, Rong Pan, Yanfang Le, Costin Raiciu, Mark Handley, Timo Schneider, Nils Blach, Ahmad Ghalayini, Daniel Alves, Michael Papamichael, Adrian Caulfield, and Torsten Hoefler. 2024. FAST-FLOW: Flexible Adaptive Congestion Control for High-Performance Datacenters. (2024). arXiv:cs.NI\/2404.01630 https:\/\/arxiv.org\/abs\/2404.01630"},{"key":"e_1_3_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1145\/224056.224076"},{"key":"e_1_3_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1145\/3651890.3672233"},{"key":"e_1_3_2_1_13_1","unstructured":"Aaron Grattafiori Abhimanyu Dubey Abhinav Jauhri et al. 2024. The Llama 3 Herd of Models. (2024). arXiv:cs.AI\/2407.21783 https:\/\/arxiv.org\/abs\/2407.21783"},{"key":"e_1_3_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1145\/3387514.3405849"},{"key":"e_1_3_2_1_15_1","volume-title":"Performance Counter Monitor. https:\/\/www.intel.com\/content\/www\/us\/en\/developer\/articles\/tool\/performance-counter-monitor.html. (Nov","author":"INTEL.","year":"2022","unstructured":"INTEL. 2022. Performance Counter Monitor. https:\/\/www.intel.com\/content\/www\/us\/en\/developer\/articles\/tool\/performance-counter-monitor.html. (Nov 2022)."},{"key":"e_1_3_2_1_16_1","doi-asserted-by":"crossref","DOI":"10.1109\/ISCAS.1998.705322","volume-title":"1998 IEEE International Symposium on Circuits and Systems (ISCAS)","volume":"6","author":"Kwon Taeck-Geun","year":"1998","unstructured":"Taeck-Geun Kwon, Sook-Hyang Lee, and June-Kyung Rho. 1998. Scheduling algorithm for real-time burst traffic using dynamic weighted round robin. In 1998 IEEE International Symposium on Circuits and Systems (ISCAS), Vol. 6. IEEE, 506\u2013509."},{"key":"e_1_3_2_1_17_1","unstructured":"Yanfang Le Rong Pan Peter Newman Jeremias Blendin Abdul Kabbani Vipin Jain Raghava Sivaramu and Francis Matus. 2024. STrack: A Reliable Multipath Transport for AI\/ML Clusters. (2024). arXiv:cs.NI\/2407.15266 https:\/\/arxiv.org\/abs\/2407.15266"},{"key":"e_1_3_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.5555\/42040.42133"},{"key":"e_1_3_2_1_19_1","volume-title":"Direct Packet Placement. https:\/\/www.freepatentsonline.com\/y2020\/0092229.html. (24","author":"David Dotan","year":"2019","unstructured":"Levi, Dotan David, Urman, and Avi. 2019. Direct Packet Placement. https:\/\/www.freepatentsonline.com\/y2020\/0092229.html. (24 Nov 2019)."},{"key":"e_1_3_2_1_20_1","volume-title":"RunD: A Lightweight Secure Container Runtime for High-density Deployment and High-concurrency Startup in Serverless Computing. In 2022 USENIX Annual Technical Conference (USENIX ATC 22)","author":"Li Zijun","year":"2022","unstructured":"Zijun Li, Jiagan Cheng, Quan Chen, Eryu Guan, Zizheng Bian, Yi Tao, Bin Zha, Qiang Wang, Weidong Han, and Minyi Guo. 2022. RunD: A Lightweight Secure Container Runtime for High-density Deployment and High-concurrency Startup in Serverless Computing. In 2022 USENIX Annual Technical Conference (USENIX ATC 22). USENIX Association, Carlsbad, CA, 53\u201368. https:\/\/www.usenix.org\/conference\/atc22\/presentation\/li-zijun-rund"},{"key":"e_1_3_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.5555\/3307441.3307472"},{"key":"e_1_3_2_1_22_1","volume-title":"15th USENIX Symposium on Networked Systems Design and Implementation (NSDI 18)","author":"Lu Yuanwei","year":"2018","unstructured":"Yuanwei Lu, Guo Chen, Bojie Li, Kun Tan, Yongqiang Xiong, Peng Cheng, Jiansong Zhang, Enhong Chen, and Thomas Moscibroda. 2018. Multi-Path Transport for RDMA in Datacenters. https:\/\/www.usenix.org\/conference\/nsdi18\/presentation\/lu. In 15th USENIX Symposium on Networked Systems Design and Implementation (NSDI 18). USENIX Association, Renton, WA, 357\u2013371."},{"key":"e_1_3_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.1145\/1810085.1810091"},{"key":"e_1_3_2_1_24_1","volume-title":"d.]. Megatron-LM. https:\/\/github.com\/NVIDIA\/Megatron-LM\/tree\/main\/megatron. ([n. d.])","author":"NVIDIA.","year":"2024","unstructured":"NVIDIA. [n. d.]. Megatron-LM. https:\/\/github.com\/NVIDIA\/Megatron-LM\/tree\/main\/megatron. ([n. d.]). Accessed: Sep 2024."},{"key":"e_1_3_2_1_25_1","volume-title":"https:\/\/docs.nvidia.com\/nvidia-mellanox-neo-documentation.pdf. (Nov","author":"Mellanox Neo NVIDIA.","year":"2023","unstructured":"NVIDIA. 2023. Mellanox Neo. https:\/\/docs.nvidia.com\/nvidia-mellanox-neo-documentation.pdf. (Nov 2023)."},{"key":"e_1_3_2_1_26_1","doi-asserted-by":"publisher","DOI":"10.1145\/2817817.2731200"},{"key":"e_1_3_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1145\/3651890.3672265"},{"key":"e_1_3_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"e_1_3_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10586-007-0046-3"},{"key":"e_1_3_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1145\/3603269.3604849"},{"key":"e_1_3_2_1_31_1","volume-title":"2020 USENIX Annual Technical Conference (USENIX ATC 20)","author":"Tian Kun","year":"2020","unstructured":"Kun Tian, Yu Zhang, Luwei Kang, Yan Zhao, and Yaozu Dong. 2020. coIOMMU: A Virtual IOMMU with Cooperative DMA Buffer Tracking for Efficient Memory Management in Direct I\/O. In 2020 USENIX Annual Technical Conference (USENIX ATC 20). USENIX Association, 479\u2013492. https:\/\/www.usenix.org\/conference\/atc20\/presentation\/tian"},{"key":"e_1_3_2_1_32_1","volume-title":"Tsirkin and Cornelia Huck","author":"Michael","year":"2023","unstructured":"Michael S. Tsirkin and Cornelia Huck. 2023. Virtual I\/O Device (VIRTIO) Version 1.3. https:\/\/docs.oasis-open.org\/virtio\/virtio\/v1.3\/virtio-v1.3.html. (Oct 2023)."},{"key":"e_1_3_2_1_33_1","volume-title":"Let It Flow: Resilient Asymmetric Load Balancing with Flowlet Switching. In 14th USENIX Symposium on Networked Systems Design and Implementation (NSDI 17)","author":"Vanini Erico","year":"2017","unstructured":"Erico Vanini, Rong Pan, Mohammad Alizadeh, Parvin Taheri, and Tom Edsall. 2017. Let It Flow: Resilient Asymmetric Load Balancing with Flowlet Switching. In 14th USENIX Symposium on Networked Systems Design and Implementation (NSDI 17). USENIX Association, Boston, MA, 407\u2013420. https:\/\/www.usenix.org\/conference\/nsdi17\/technical-sessions\/presentation\/vanini"}],"event":{"name":"SIGCOMM '25: ACM SIGCOMM 2025 Conference","location":"S\u00e3o Francisco Convent Coimbra Portugal","acronym":"SIGCOMM '25","sponsor":["SIGCOMM ACM Special Interest Group on Data Communication"]},"container-title":["Proceedings of the ACM SIGCOMM 2025 Conference"],"original-title":[],"link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3718958.3750539","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,27]],"date-time":"2025-08-27T16:54:17Z","timestamp":1756313657000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3718958.3750539"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,27]]},"references-count":33,"alternative-id":["10.1145\/3718958.3750539","10.1145\/3718958"],"URL":"https:\/\/doi.org\/10.1145\/3718958.3750539","relation":{},"subject":[],"published":{"date-parts":[[2025,8,27]]},"assertion":[{"value":"2025-08-27","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}