{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,21]],"date-time":"2026-01-21T07:32:18Z","timestamp":1768980738563,"version":"3.49.0"},"publisher-location":"New York, NY, USA","reference-count":40,"publisher":"ACM","license":[{"start":{"date-parts":[[2024,4,22]],"date-time":"2024-04-22T00:00:00Z","timestamp":1713744000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-sa\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100006374","name":"European Research Council","doi-asserted-by":"publisher","award":["770889"],"award-info":[{"award-number":["770889"]}],"id":[{"id":"10.13039\/501100006374","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2024,4,22]]},"DOI":"10.1145\/3642968.3654820","type":"proceedings-article","created":{"date-parts":[[2024,4,17]],"date-time":"2024-04-17T00:04:51Z","timestamp":1713312291000},"page":"43-48","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Toward GPU-centric Networking on Commodity Hardware"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9400-324X","authenticated-orcid":false,"given":"Massimo","family":"Girondi","sequence":"first","affiliation":[{"name":"KTH Royal Institute of Technology, Stockholm, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9780-873X","authenticated-orcid":false,"given":"Mariano","family":"Scazzariello","sequence":"additional","affiliation":[{"name":"KTH Royal Institute of Technology, Stockholm, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6066-746X","authenticated-orcid":false,"suffix":"Jr","given":"Gerald Q.","family":"Maguire","sequence":"additional","affiliation":[{"name":"KTH Royal Institute of Technology, Stockholm, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1256-1070","authenticated-orcid":false,"given":"Dejan","family":"Kosti\u0107","sequence":"additional","affiliation":[{"name":"KTH Royal Institute of Technology, Stockholm, Sweden"}]}],"member":"320","published-online":{"date-parts":[[2024,4,22]]},"reference":[{"key":"e_1_3_2_1_1_1","unstructured":"[n. d.]. RDMA core userspace library. https:\/\/github.com\/linux-rdma\/rdma-core"},{"key":"e_1_3_2_1_2_1","unstructured":"Elena Agostini et al. 2023. Realizing the Power of Real-Time Network Processing with NVIDIA DOCA GPUNetIO. https:\/\/developer.nvidia.com\/blog\/realizing-the-power-of-real-time-network-processing-with- nvidia-doca-gpunetio\/"},{"key":"e_1_3_2_1_3_1","unstructured":"Guido Appenzeller et al. 2023. Navigating the High Cost of AI Compute. https:\/\/a16z.com\/navigating-the-high-cost-of-ai-compute\/"},{"key":"e_1_3_2_1_4_1","doi-asserted-by":"crossref","unstructured":"Luiz Barroso et al. 2017. Attack of the Killer Microseconds. Commun. ACM 60 4 (2017).","DOI":"10.1145\/3015146"},{"key":"e_1_3_2_1_5_1","unstructured":"Alexis Bjorlin. 2022. Infrastructure for Large Scale AI: \"Empowering Open\". https:\/\/drive.google.com\/file\/d\/1qqjo-5JtYAcRlK_LWYuQFH-b9MoFOP02\/view"},{"key":"e_1_3_2_1_6_1","volume-title":"NVIDIA Hopper H100 GPU: Scaling Performance","author":"Choquette Jack","year":"2023","unstructured":"Jack Choquette. 2023. NVIDIA Hopper H100 GPU: Scaling Performance. IEEE Micro 43, 3 (2023)."},{"key":"e_1_3_2_1_7_1","unstructured":"Microsoft Corporation. 2022. Microsoft Collective Communication Library. https:\/\/github.com\/microsoft\/msccl"},{"key":"e_1_3_2_1_8_1","unstructured":"NVIDIA Corporation. [n. d.]. NVIDIA Triton Inference Server. https:\/\/www.nvidia.com\/en-us\/ai-data-science\/products\/triton-inference-server\/"},{"key":"e_1_3_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1145\/2931088.2931091"},{"key":"e_1_3_2_1_10_1","unstructured":"Chris Dickson. [n. d.]. The Technology Behind A Low Latency Cloud Gaming Service. https:\/\/parsec.app\/blog\/description-of-parsec-technology-b2738dcc3842"},{"key":"e_1_3_2_1_11_1","unstructured":"The Apache Software Foundation. [n.d.]. Apache TVM. https:\/\/tvm.apache.org\/"},{"key":"e_1_3_2_1_12_1","unstructured":"Aishwarya Goel. 2023. Unraveling GPU Inference Costs for Fine-tuned Open-Source Models V\/S Closed Platforms. https:\/\/mlops.community\/unraveling-gpu-inference-costs-for-fine-tuned-open-source-models-v-s-closed-platforms\/"},{"key":"e_1_3_2_1_13_1","volume-title":"OSDI","author":"Gujarati Arpan","year":"2020","unstructured":"Arpan Gujarati, et al. 2020. Serving DNNs like Clockwork: Performance Predictability from the Bottom Up. In OSDI 2020. USA, Article 25."},{"key":"e_1_3_2_1_14_1","doi-asserted-by":"crossref","unstructured":"Yashuang Guo et al. 2020. Adaptive Bitrate Streaming in Wireless Networks With Transcoding at Network Edge Using Deep Reinforcement Learning. IEEE TVT 69 4 (2020).","DOI":"10.1109\/TVT.2020.2968498"},{"key":"e_1_3_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA.2018.00059"},{"key":"e_1_3_2_1_16_1","unstructured":"InfiniBand\u2122Trade Association. 2010. RDMA over Converged Ethernet (RoCE). https:\/\/web.archive.org\/web\/20160309123709\/https:\/\/cw.infinibandta.org\/document\/dl\/7148"},{"key":"e_1_3_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM42981.2021.9488699"},{"key":"e_1_3_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1145\/3578338.3593571"},{"key":"e_1_3_2_1_19_1","first-page":"1","article-title":"Evaluating modern GPU interconnect: PCIe, NVLink, NV-SLI, NVSwitch and GPUDirect","volume":"31","author":"Li Ang","year":"2019","unstructured":"Ang Li, et al. 2019. Evaluating modern GPU interconnect: PCIe, NVLink, NV-SLI, NVSwitch and GPUDirect. TPDS 31, 1 (July 2019).","journal-title":"TPDS"},{"key":"e_1_3_2_1_20_1","volume-title":"SIGGRAPH","author":"Pietro","year":"2019","unstructured":"Pietro Lungaro et al. 2019. Immersivemote: combining foveated AI and streaming for immersive remote operations. In SIGGRAPH 2019. Article 17."},{"key":"e_1_3_2_1_21_1","volume-title":"CSUR 2015 47","author":"Sparsh","year":"2015","unstructured":"Sparsh Mittal et al. 2015. A Survey of CPU-GPU Heterogeneous Computing Techniques. CSUR 2015 47 (2015)."},{"key":"e_1_3_2_1_22_1","unstructured":"MLCommons. 2023. Inference Datacenter v3.0 Results. https:\/\/mlcommons.org\/en\/inference-datacenter-30\/"},{"key":"e_1_3_2_1_23_1","unstructured":"Timothy Prickett Morgan. 2023. Meta Platforms Is Determined To Make Ethernet Work For AI. https:\/\/www.nextplatform.com\/2023\/09\/26\/meta-platforms-is-determined-to-make-ethernet-work-for-ai\/"},{"key":"e_1_3_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1145\/1365490.1365500"},{"key":"e_1_3_2_1_25_1","unstructured":"NVIDIA. 2023. GPUDirect Documentation. https:\/\/docs.nvidia.com\/cuda\/gpudirect-rdma\/index.html"},{"key":"e_1_3_2_1_26_1","unstructured":"NVIDIA. 2023. NVIDIA NCCL. https:\/\/developer.nvidia.com\/nccl"},{"key":"e_1_3_2_1_27_1","unstructured":"NVIDIA Corporation. 2023. Device Graph Launch. https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/index.html#device-graph-launch"},{"key":"e_1_3_2_1_28_1","unstructured":"NVIDIA Corporation. 2023. NVIDIA H100 GPU DataSheet. https:\/\/resources.nvidia.com\/en-us-tensor-core\/nvidia-tensor-core-gpu-datasheet"},{"key":"e_1_3_2_1_29_1","unstructured":"NVIDIA Corporation. 2024. Unified Communication - X Framework. https:\/\/docs.nvidia.com\/networking\/display\/HPCXv215\/Unified+Communication+-+X+Framework+Library"},{"key":"e_1_3_2_1_30_1","unstructured":"OpenAI. 2024. SORA. https:\/\/openai.com\/sora\/"},{"key":"e_1_3_2_1_31_1","doi-asserted-by":"crossref","unstructured":"Kevin R\u00f6bert et al. 2023. Latency-Aware Scheduling for Real-Time Application Support in Edge Computing. In EdgeSys 2023.","DOI":"10.1145\/3578354.3592866"},{"key":"e_1_3_2_1_32_1","volume-title":"INFaaS: Automated Model-less Inference Serving. In USENIX ATC","author":"Romero Francisco","year":"2021","unstructured":"Francisco Romero, et al. 2021. INFaaS: Automated Model-less Inference Serving. In USENIX ATC 2021."},{"key":"e_1_3_2_1_33_1","unstructured":"Eurovision Services. 2021. Eurovision Services successfully tests full cloud production solution. https:\/\/www.eurovision.net\/insights\/technical\/eurovision-services-successfully-tests-full-cloud-production-solution"},{"key":"e_1_3_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.207"},{"key":"e_1_3_2_1_35_1","volume-title":"PFC: Detailed Evaluation. In KBNets","author":"Shpiner Alexander","year":"2017","unstructured":"Alexander Shpiner, et al. 2017. RoCE Rocks without PFC: Detailed Evaluation. In KBNets 2017."},{"key":"e_1_3_2_1_36_1","doi-asserted-by":"crossref","unstructured":"Mark Silberstein et al. 2016. GP Unet: Networking abstractions for GPU programs. TOCS 34 3 (2016).","DOI":"10.1145\/2963098"},{"key":"e_1_3_2_1_37_1","doi-asserted-by":"publisher","DOI":"10.1145\/3373376.3378528"},{"key":"e_1_3_2_1_38_1","volume-title":"Handbook of Fiber Optic Data Communication","author":"Wadekar Manoj","unstructured":"Manoj Wadekar. 2013. Chapter 11 - InfiniBand, iWARP, and RoCE. In Handbook of Fiber Optic Data Communication (Fourth Edition) (fourth ed.), Casimer DeCusatis (Ed.). Academic Press, Oxford."},{"key":"e_1_3_2_1_39_1","unstructured":"Matt Walsh. 2023. ChatGPT Statistics (2023) --- The Key Facts and Figures. https:\/\/www.stylefactoryproductions.com\/blog\/chatgpt-statistics"},{"key":"e_1_3_2_1_40_1","volume-title":"SHEPHERD: Serving DNNs in the Wild. In NSDI","author":"Zhang Hong","year":"2023","unstructured":"Hong Zhang, et al. 2023. SHEPHERD: Serving DNNs in the Wild. In NSDI 2023."}],"event":{"name":"EuroSys '24: Nineteenth European Conference on Computer Systems","location":"Athens Greece","acronym":"EuroSys '24","sponsor":["SIGOPS ACM Special Interest Group on Operating Systems"]},"container-title":["Proceedings of the 7th International Workshop on Edge Systems, Analytics and Networking"],"original-title":[],"link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3642968.3654820","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3642968.3654820","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,23]],"date-time":"2025-08-23T18:43:08Z","timestamp":1755974588000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3642968.3654820"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,4,22]]},"references-count":40,"alternative-id":["10.1145\/3642968.3654820","10.1145\/3642968"],"URL":"https:\/\/doi.org\/10.1145\/3642968.3654820","relation":{},"subject":[],"published":{"date-parts":[[2024,4,22]]},"assertion":[{"value":"2024-04-22","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}