{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,1]],"date-time":"2025-11-01T05:56:52Z","timestamp":1761976612747,"version":"build-2065373602"},"reference-count":44,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T00:00:00Z","timestamp":1755043200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T00:00:00Z","timestamp":1755043200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,8,13]]},"DOI":"10.1109\/hpcc67675.2025.00092","type":"proceedings-article","created":{"date-parts":[[2025,10,31]],"date-time":"2025-10-31T17:09:09Z","timestamp":1761930549000},"page":"584-591","source":"Crossref","is-referenced-by-count":0,"title":["APSCC: Adaptive Congestion Control for Packet-Sprayed RDMA Networks in AI Clusters"],"prefix":"10.1109","author":[{"given":"Jian","family":"Tang","sequence":"first","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wenming","family":"Zheng","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fangfang","family":"Yan","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaoping","family":"Fan","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Luren","family":"Liu","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anran","family":"Xu","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yi","family":"Liao","sequence":"additional","affiliation":[{"name":"China Telecom"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"journal-title":"Deepseek-v3 technical report","year":"2024","author":"Liu","key":"ref2"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM42981.2021.9488875"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM53939.2023.10228968"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/CLUSTER.2019.8891004"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3544216.3544238"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/2934872.2934908"},{"key":"ref8","first-page":"49","article-title":"Empowering Azure Storage with RDMA","volume-title":"Proceedings of USENIX NSDI","author":"Bai","year":"2023"},{"volume-title":"Nvidia Collective Communications Library (NCCL)","key":"ref9"},{"key":"ref10","article-title":"When cloud storage meets RDMA","volume-title":"Proceedings of NSDI","author":"Gao","year":"2021"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3651890.3672233"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3651890.3672265"},{"journal-title":"STrack: A Reliable Multipath Transport for AI\/ML Clusters","year":"2024","author":"Le","key":"ref13"},{"volume-title":"To Spray or Not to Spray","author":"Shokarev","key":"ref14"},{"volume-title":"Adaptive Routing","key":"ref15"},{"key":"ref16","first-page":"407","article-title":"Let it flow: Resilient asymmetric load balancing with flowlet switching","volume-title":"Proceedings of USENIX NSDI","author":"Vanini","year":"2017"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/INFCOM.2013.6567015"},{"journal-title":"Challenging the need for packet spraying in large-scale distributed training","year":"2024","author":"Addanki","key":"ref18"},{"journal-title":"Reps: Recycled entropy packet spraying for adaptive load balancing and failure mitigation","year":"2024","author":"Bonato","key":"ref19"},{"volume-title":"Network Simulator 3","year":"2020","key":"ref20"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1145\/2785956.2787510"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3387514.3406591"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1145\/2785956.2787484"},{"key":"ref24","first-page":"19","article-title":"Re-architecting congestion management in lossless Ethernet","volume-title":"Proceedings of NSDI","author":"Cheng","year":"2020"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ICNP55882.2022.9940420"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2021.3091475"},{"key":"ref27","article-title":"Simai: Unifying architecture design and performance tunning for large-scale large language model training with scalability and precision","volume-title":"Proceedings of NSDI","author":"Wang","year":"2025"},{"journal-title":"Deep gradient compression: Reducing the communication bandwidth for distributed training","year":"2017","author":"Lin","key":"ref28"},{"volume-title":"Spectrumx","key":"ref29"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/CCGRID.2017.102"},{"key":"ref31","first-page":"219","article-title":"Bolt:Sub-RTT Congestion Control for Ultra-Low Latency","volume-title":"Proceedings of USENIX NSDI","author":"Arslan","year":"2023"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/3386367.3431316"},{"volume-title":"ZTRCC","key":"ref33"},{"volume-title":"Programmable Congestion Control (PCC)","key":"ref34"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/WOWMOM.2009.5282435"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-39867-7_26"},{"key":"ref37","first-page":"1","article-title":"SRNIC: A Scalable Architecture for RDMA NICs","volume-title":"Proceedings of USENIX NSDI","author":"Wang","year":"2023"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/2999572.2999593"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1145\/1851182.1851192"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1145\/2413176.2413181"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1145\/3452296.3472927"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1145\/3341302.3342085"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/90.251892"},{"journal-title":"Infiniband architecture volume 1, general specifications, release 1.4","year":"2020","key":"ref44"}],"event":{"name":"2025 IEEE International Conference on High Performance Computing and Communications (HPCC)","start":{"date-parts":[[2025,8,13]]},"location":"Exeter, United Kingdom","end":{"date-parts":[[2025,8,15]]}},"container-title":["2025 IEEE International Conference on High Performance Computing and Communications (HPCC)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11207229\/11207226\/11207411.pdf?arnumber=11207411","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,1]],"date-time":"2025-11-01T05:52:30Z","timestamp":1761976350000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11207411\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,13]]},"references-count":44,"URL":"https:\/\/doi.org\/10.1109\/hpcc67675.2025.00092","relation":{},"subject":[],"published":{"date-parts":[[2025,8,13]]}}}