{"id":"https://openalex.org/W4414079529","doi":"https://doi.org/10.1109/tvlsi.2025.3604705","title":"MoE-Sched: Enabling Efficient FPGA Deployment of Mixture-of-Experts Vision Transformers via Coordinated Scheduling","display_name":"MoE-Sched: Enabling Efficient FPGA Deployment of Mixture-of-Experts Vision Transformers via Coordinated Scheduling","publication_year":2025,"publication_date":"2025-09-08","ids":{"openalex":"https://openalex.org/W4414079529","doi":"https://doi.org/10.1109/tvlsi.2025.3604705"},"language":"en","primary_location":{"id":"doi:10.1109/tvlsi.2025.3604705","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvlsi.2025.3604705","pdf_url":null,"source":{"id":"https://openalex.org/S37538908","display_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","issn_l":"1063-8210","issn":["1063-8210","1557-9999"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102379923","display_name":"Jiale Dong","orcid":null},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiale Dong","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0009-0006-0188-1681","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034765006","display_name":"Wenqi Lou","orcid":"https://orcid.org/0000-0002-2240-6672"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenqi Lou","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0000-0002-2240-6672","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102952321","display_name":"Hao Wu","orcid":"https://orcid.org/0009-0005-2207-737X"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hao Wu","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104095639","display_name":"Zhendong Zheng","orcid":"https://orcid.org/0000-0002-6486-3062"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhendong Zheng","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014908428","display_name":"Yunji Qin","orcid":"https://orcid.org/0009-0005-9404-581X"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yunji Qin","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072915437","display_name":"Lei Gong","orcid":"https://orcid.org/0000-0002-8391-5526"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lei Gong","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0000-0002-8391-5526","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011884768","display_name":"Chao Wang","orcid":"https://orcid.org/0000-0002-9403-5575"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chao Wang","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0000-0002-9403-5575","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5077322091","display_name":"Xuehai Zhou","orcid":"https://orcid.org/0000-0002-8360-3143"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuehai Zhou","raw_affiliation_strings":["School of Computer Science, University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0000-0002-8360-3143","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I126520041"],"apc_list":null,"apc_paid":null,"fwci":0.9253,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.77225381,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":"34","issue":"1","first_page":"104","last_page":"117"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.9955999851226807,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.9955999851226807,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9700000286102295,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10191","display_name":"Robotics and Sensor-Based Localization","score":0.9366000294685364,"subfield":{"id":"https://openalex.org/subfields/2202","display_name":"Aerospace Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.757099986076355},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6360999941825867},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.6100000143051147},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.569100022315979},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.5511999726295471},{"id":"https://openalex.org/keywords/efficient-energy-use","display_name":"Efficient energy use","score":0.45239999890327454}],"concepts":[{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.757099986076355},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7143999934196472},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6360999941825867},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.6100000143051147},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.569100022315979},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5511999726295471},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.45559999346733093},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.45239999890327454},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.4462999999523163},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.41519999504089355},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.387800008058548},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.35109999775886536},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.32820001244544983},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.31139999628067017},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.29330000281333923},{"id":"https://openalex.org/C58013763","wikidata":"https://www.wikidata.org/wiki/Q5754574","display_name":"High-level synthesis","level":3,"score":0.2800999879837036},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.2786000072956085},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.2558000087738037},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tvlsi.2025.3604705","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvlsi.2025.3604705","pdf_url":null,"source":{"id":"https://openalex.org/S37538908","display_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","issn_l":"1063-8210","issn":["1063-8210","1557-9999"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G461544363","display_name":null,"funder_award_id":"BK20241818","funder_id":"https://openalex.org/F4320322769","funder_display_name":"Natural Science Foundation of Jiangsu Province"},{"id":"https://openalex.org/G6472083938","display_name":null,"funder_award_id":"62172380","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7984994750","display_name":null,"funder_award_id":"62502489","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320322769","display_name":"Natural Science Foundation of Jiangsu Province","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W2795961274","https://openalex.org/W2917087921","https://openalex.org/W3138516171","https://openalex.org/W3198625877","https://openalex.org/W3206453033","https://openalex.org/W4283815817","https://openalex.org/W4285601701","https://openalex.org/W4286001027","https://openalex.org/W4308479898","https://openalex.org/W4309125142","https://openalex.org/W4312677967","https://openalex.org/W4360831786","https://openalex.org/W4375868854","https://openalex.org/W4379115894","https://openalex.org/W4386076539","https://openalex.org/W4387789586","https://openalex.org/W4389162698","https://openalex.org/W4389337739","https://openalex.org/W4390873360","https://openalex.org/W4390873673","https://openalex.org/W4393141090","https://openalex.org/W4399487406","https://openalex.org/W4400524700","https://openalex.org/W4402727538","https://openalex.org/W4404133759","https://openalex.org/W4406893597","https://openalex.org/W4409282814","https://openalex.org/W4411725841"],"related_works":["https://openalex.org/W2770234245","https://openalex.org/W96612179","https://openalex.org/W4229499248","https://openalex.org/W2566006169","https://openalex.org/W1567818861","https://openalex.org/W2987774938","https://openalex.org/W4256492088","https://openalex.org/W1967938402","https://openalex.org/W2386041993","https://openalex.org/W1608572506"],"abstract_inverted_index":{"Compared":[0],"to":[1,11,108,114,174],"traditional":[2],"Vision":[3],"Transformers":[4],"(ViT),":[5],"Mixture-of-Experts":[6],"ViTs":[7],"(MoE-ViTs)":[8],"are":[9],"introduced":[10],"scale":[12],"model":[13],"size":[14],"without":[15],"a":[16,24,85,91,126,157],"proportional":[17],"increase":[18],"in":[19,65,77,166],"computational":[20],"complexity,":[21],"making":[22],"them":[23],"new":[25],"research":[26],"focus.":[27],"Given":[28],"the":[29,57,75],"high":[30],"performance":[31],"and":[32,60,71,111,120,134,168],"reconfigurability,":[33],"field-programmable":[34],"gate":[35],"array":[36],"(FPGA)-based":[37],"accelerators":[38,50],"for":[39,131],"MoE-ViTs":[40,79],"emerge,":[41],"delivering":[42],"substantial":[43],"gains":[44],"over":[45,169],"general-purpose":[46],"processors.":[47],"However,":[48],"existing":[49],"often":[51],"fall":[52],"short":[53],"of":[54],"efficiently":[55],"managing":[56],"highly":[58],"dynamic":[59],"sparse":[61,135],"computation":[62],"patterns,":[63],"resulting":[64],"suboptimal":[66],"tradeoffs":[67],"between":[68],"resource":[69,121],"utilization":[70],"performance.":[72],"To":[73],"address":[74],"inefficiencies":[76],"deploying":[78],"on":[80,153],"FPGAs,":[81],"we":[82],"present":[83],"MoE-Sched,":[84],"novel":[86],"end-to-end":[87],"accelerator":[88,148],"that":[89,146],"embraces":[90],"scheduling-centric":[92],"design":[93],"philosophy.":[94],"Rather":[95],"than":[96,183],"optimizing":[97],"isolated":[98],"kernels,":[99],"MoE-Sched":[100],"coordinates":[101],"multilevel":[102],"scheduling,":[103],"from":[104],"fine-grained":[105],"intrakernel":[106],"streaming":[107,132],"module":[109],"reuse":[110],"multidie":[112],"mapping,":[113],"holistically":[115],"balance":[116],"latency,":[117],"bandwidth":[118],"(BW),":[119],"usage.":[122],"We":[123],"further":[124],"integrate":[125],"hardware-aware":[127],"quantization":[128],"scheme":[129],"tailored":[130],"attention":[133],"expert":[136],"execution,":[137],"preserving":[138],"accuracy":[139,185],"while":[140,180],"minimizing":[141],"overhead.":[142],"Experimental":[143],"results":[144],"demonstrate":[145],"our":[147],"achieves":[149],"nearly":[150],"100":[151],"frames/s":[152],"M<sup":[154],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[155,159],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">3</sup>ViT-tiny,":[156],"<inline-formula":[158],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">":[160],"<tex-math":[161],"notation=\"LaTeX\">$3.13\\times":[162],"$</tex-math>":[163],"</inline-formula>":[164],"improvement":[165],"throughput,":[167],"75%":[170],"energy":[171],"reduction":[172],"compared":[173],"state-of-the-art":[175],"(SOTA)":[176],"FPGA":[177],"MoE":[178],"accelerators,":[179],"maintaining":[181],"less":[182],"1%":[184],"loss":[186],"across":[187],"vision":[188],"benchmarks.":[189],"Our":[190],"implementation":[191],"will":[192],"be":[193],"open-sourced.":[194]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2025-12-30T23:08:21.542490","created_date":"2025-10-10T00:00:00"}
