{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,8]],"date-time":"2026-03-08T01:54:31Z","timestamp":1772934871682,"version":"3.50.1"},"reference-count":40,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T00:00:00Z","timestamp":1765152000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T00:00:00Z","timestamp":1765152000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100000001","name":"NSF","doi-asserted-by":"publisher","award":["2211018,2116962,1931531"],"award-info":[{"award-number":["2211018,2116962,1931531"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,8]]},"DOI":"10.1109\/bigdata66926.2025.11401442","type":"proceedings-article","created":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T20:57:57Z","timestamp":1772830677000},"page":"201-210","source":"Crossref","is-referenced-by-count":0,"title":["FLEXI: Phase-Aware Function Resizing for Heterogeneous Serverless GPU Workloads"],"prefix":"10.1109","author":[{"given":"Shruti","family":"Mohanty","sequence":"first","affiliation":[{"name":"Computer Science and Engineering, Pennsylvania State University"}]},{"given":"Vivek M.","family":"Bhasi","sequence":"additional","affiliation":[{"name":"Computer Science and Engineering, Pennsylvania State University"}]},{"given":"Jashwant R.","family":"Gunasekaran","sequence":"additional","affiliation":[{"name":"Adobe Research"}]},{"given":"Prashanth","family":"Thinakaran","sequence":"additional","affiliation":[{"name":"Clockwork Systems"}]},{"given":"Mahmut T.","family":"Kandemir","sequence":"additional","affiliation":[{"name":"Computer Science and Engineering, Pennsylvania State University"}]},{"given":"Chita R.","family":"Das","sequence":"additional","affiliation":[{"name":"Computer Science and Engineering, Pennsylvania State University"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/3698038.3698560"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1145\/3588195.3592988"},{"key":"ref3","article-title":"Gpu as a service (gpuaas) market report","volume-title":"Grand View Research","year":"2024"},{"key":"ref4","volume-title":"Serverless gpu overview: Unlocking ai innovation and productivity in hpc tasks","year":"2024"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS57955.2024.00018"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507709"},{"key":"ref7","first-page":"42","article-title":"Esg: Cost-efficient inference scheduling on heterogeneous clusters","author":"Hui","year":"2024","journal-title":"HPDC"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575721"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507732"},{"key":"ref10","doi-asserted-by":"crossref","first-page":"32","DOI":"10.1016\/j.jpdc.2020.01.004","article-title":"Accelerated serverless computing based on gpu virtualization","volume":"139","author":"Naranjo","year":"2020","journal-title":"Journal of Parallel and Distributed Computing"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS53621.2022.00077"},{"key":"ref12","volume-title":"Using serverless gpus in azure container apps","year":"2025"},{"key":"ref13","volume-title":"Cloud run gpus, now ga, makes running ai workloads easier for everyone","year":"2025"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00073"},{"key":"ref15","first-page":"1049","article-title":"{MArk}: Exploiting cloud services for {Cost-Effective},{SLO-Aware} machine learning inference serving","author":"Zhang","year":"2019","journal-title":"USENIX ATC 19"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3652892.3700760"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/3542929.3563510"},{"key":"ref18","first-page":"199","article-title":"Serving heterogeneous machine learning models on Multi-GPU servers with Spatio-Temporal sharing","author":"Choi","year":"2022","journal-title":"USENIX ATC"},{"key":"ref19","first-page":"311","article-title":"Dilu: Enabling gpu resourcing-on-demand for serverless dl serving via introspective elasticity","author":"Zhang","year":"2025","journal-title":"ACM ASPLOS"},{"key":"ref20","article-title":"Has-gpu: Hybrid auto-scaling with fine-grained gpu allocation for serverless dl inference","author":"Li","year":"2025","journal-title":"arXiv preprint"},{"key":"ref21","first-page":"995","article-title":"Beware of fragmentation: Scheduling {GPU-Sharing} workloads with fragmentation gradient descent","author":"Weng","year":"2023","journal-title":"USENIX ATC"},{"key":"ref22","first-page":"427","article-title":"Pocket: Elastic ephemeral storage for serverless analytics","author":"Klimovic","year":"2018","journal-title":"USENIX OSDI"},{"key":"ref23","first-page":"285","article-title":"{SONIC}: Application-aware data passing for chained serverless applications","author":"Mahgoub","year":"2021","journal-title":"USENIX ATC"},{"key":"ref24","author":"Yu","year":"2025","journal-title":"Torpor: Gpu-enabled serverless computing for low-latency, resource-efficient inference"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3544216.3544224"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/SIES62473.2024.10767927"},{"key":"ref27","first-page":"205","article-title":"Serverless in the wild: Characterizing and optimizing the serverless workload at a large cloud provider","author":"Shahrad","year":"2020","journal-title":"USENIX ATC 20"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/3620678.3624783"},{"key":"ref29","volume-title":"Azure Public Dataset","year":"2023"},{"key":"ref30","volume-title":"AWS Lambda","year":"2023"},{"key":"ref31","volume-title":"Microsoft Azure Serverless Functions","year":"2023"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref33","author":"Sanh","year":"2019","journal-title":"Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter"},{"key":"ref34","volume-title":"Mistral 7b","author":"Jiang","year":"2023"},{"issue":"7","key":"ref35","article-title":"Cupy: A numpy-compatible library for nvidia gpu calculations","volume":"151","author":"Nishino","year":"2017","journal-title":"NeurIPS"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1038\/nbt.3988"},{"key":"ref37","first-page":"945","article-title":"MLaaS in the wild: Workload analysis and scheduling in Large-Scale heterogeneous GPU clusters","author":"Weng","year":"2022","journal-title":"USENIX NSDI"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/3552326.3567506"},{"key":"ref39","article-title":"Jiagu: Adaptive serverless autoscaling via predictive and elastic allocation","author":"Liu","year":"2024","journal-title":"USENIX ATC"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS57875.2023.00023"}],"event":{"name":"2025 IEEE International Conference on Big Data (BigData)","location":"Macau, China","start":{"date-parts":[[2025,12,8]]},"end":{"date-parts":[[2025,12,11]]}},"container-title":["2025 IEEE International Conference on Big Data (BigData)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11400704\/11400712\/11401442.pdf?arnumber=11401442","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T07:19:05Z","timestamp":1772867945000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11401442\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,8]]},"references-count":40,"URL":"https:\/\/doi.org\/10.1109\/bigdata66926.2025.11401442","relation":{},"subject":[],"published":{"date-parts":[[2025,12,8]]}}}