{"id":"https://openalex.org/W7214729152","doi":"https://doi.org/10.48550/arxiv.2609.33207","title":"MorphAtt: A Neuromorphic Accelerator for Efficient Multi-Head Attention Processing in Spiking Vision Transformers","display_name":"MorphAtt: A Neuromorphic Accelerator for Efficient Multi-Head Attention Processing in Spiking Vision Transformers","publication_year":2026,"publication_date":"2026-09-27","ids":{"openalex":"https://openalex.org/W7214729152","doi":"https://doi.org/10.48550/arxiv.2609.33207"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.33207","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33207","pdf_url":"https://arxiv.org/pdf/2609.33207","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.33207","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5030422145","display_name":"Rachmad Vidya Wicaksana Putra","orcid":"https://orcid.org/0000-0001-8597-4530","observed_orcids":["https://orcid.org/0000-0001-8597-4530"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Putra, Rachmad Vidya Wicaksana","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153503750","display_name":"Amirhesam Jafari Rad","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rad, Amirhesam Jafari","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134142884","display_name":"Muhammad Shafique","orcid":"https://orcid.org/0000-0002-2607-8135","observed_orcids":["https://orcid.org/0000-0002-2607-8135"]},"institutions":[{"id":"https://openalex.org/I120250893","display_name":"New York University Abu Dhabi","ror":"https://ror.org/00e5k0821","country_code":"AE","type":"education","lineage":["https://openalex.org/I120250893","https://openalex.org/I57206974"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Shafique, Muhammad","raw_affiliation_strings":["eBRAIN Lab , New York University (NYU) Abu Dhabi , Abu Dhabi , United Arab Emirates"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"eBRAIN Lab , New York University (NYU) Abu Dhabi , Abu Dhabi , United Arab Emirates","institution_ids":["https://openalex.org/I120250893"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.6830999851226807,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.6830999851226807,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.23749999701976776,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.05009999871253967,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/neuromorphic-accelerators","display_name":"neuromorphic accelerators","score":0.8500000238418579},{"id":"https://openalex.org/keywords/multi-head-self-attention","display_name":"multi-head self-attention","score":0.9739999771118164},{"id":"https://openalex.org/keywords/spiking-neural-networks","display_name":"spiking neural networks","score":0.6930000185966492},{"id":"https://openalex.org/keywords/energy-efficient-inference","display_name":"energy-efficient inference","score":0.6000000238418579},{"id":"https://openalex.org/keywords/edge-intelligence","display_name":"edge AI","score":0.6309999823570251}],"concepts":[{"id":"https://openalex.org/C151927369","wikidata":"https://www.wikidata.org/wiki/Q1981312","display_name":"Neuromorphic engineering","level":3,"score":0.751800000667572},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7422999739646912},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.7251999974250793},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5436000227928162},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.5292999744415283},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.5133000016212463},{"id":"https://openalex.org/C118993495","wikidata":"https://www.wikidata.org/wiki/Q5042828","display_name":"Electrical efficiency","level":3,"score":0.42559999227523804},{"id":"https://openalex.org/C11731999","wikidata":"https://www.wikidata.org/wiki/Q9067355","display_name":"Spiking neural network","level":3,"score":0.4113999903202057},{"id":"https://openalex.org/C2984118289","wikidata":"https://www.wikidata.org/wiki/Q29954","display_name":"Power consumption","level":3,"score":0.40849998593330383},{"id":"https://openalex.org/C46362747","wikidata":"https://www.wikidata.org/wiki/Q173431","display_name":"CMOS","level":2,"score":0.40230000019073486},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.38769999146461487},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.36250001192092896},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3573000133037567},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3368000090122223},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3073999881744385},{"id":"https://openalex.org/C182019814","wikidata":"https://www.wikidata.org/wiki/Q1143830","display_name":"Resistive random-access memory","level":3,"score":0.30250000953674316},{"id":"https://openalex.org/C2780992000","wikidata":"https://www.wikidata.org/wiki/Q17016113","display_name":"Generator (circuit theory)","level":3,"score":0.3000999987125397},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.2969000041484833},{"id":"https://openalex.org/C84462506","wikidata":"https://www.wikidata.org/wiki/Q173142","display_name":"Digital signal processing","level":2,"score":0.28450000286102295},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.27730000019073486},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.2720000147819519},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.26660001277923584},{"id":"https://openalex.org/C2982832238","wikidata":"https://www.wikidata.org/wiki/Q5531640","display_name":"General purpose","level":2,"score":0.25929999351501465},{"id":"https://openalex.org/C80469333","wikidata":"https://www.wikidata.org/wiki/Q189088","display_name":"Von Neumann architecture","level":2,"score":0.2502000033855438}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.33207","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33207","pdf_url":"https://arxiv.org/pdf/2609.33207","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.33207","doi":"https://doi.org/10.48550/arxiv.2609.33207","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.33207","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.33207","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33207","pdf_url":"https://arxiv.org/pdf/2609.33207","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214729152.pdf","grobid_xml":"https://content.openalex.org/works/W7214729152.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7157226155","https://openalex.org/W4413469639","https://openalex.org/W7165189742","https://openalex.org/W4410949436","https://openalex.org/W7117735714","https://openalex.org/W7104990936","https://openalex.org/W4414500557","https://openalex.org/W4403785848","https://openalex.org/W7212354517","https://openalex.org/W4411206652"],"abstract_inverted_index":{"Spiking":[0,77,83],"Vision":[1],"Transformers":[2],"(SViTs)":[3],"are":[4,107],"developed":[5],"as":[6],"an":[7],"energy-efficient":[8,162],"alternative":[9],"to":[10,33,138],"conventional":[11],"ViTs":[12],"for":[13],"computer":[14],"vision":[15],"tasks":[16],"at":[17,166],"the":[18,110,167],"edge.":[19,168],"However,":[20],"huge":[21],"parameter":[22],"counts":[23],"and":[24,88,101,131,154],"complex":[25],"multi-head":[26],"self-attention":[27],"(MHSA)":[28],"operations":[29,71],"make":[30],"it":[31,68],"challenging":[32],"achieve":[34],"high":[35],"energy":[36,142],"efficiency":[37,48,155],"in":[38,42,97],"SViT":[39,51,62],"inference,":[40],"especially":[41],"tightly":[43],"constrained":[44],"applications.":[45],"To":[46,93],"maximize":[47],"gains":[49],"of":[50,122,128,134,141],"processing,":[52],"we":[53],"propose":[54],"MorphAtt,":[55],"a":[56,76,81],"novel":[57],"digital":[58],"accelerator":[59],"that":[60,148],"expedites":[61],"inference":[63],"through":[64],"streamlined":[65],"processing.":[66],"Specifically,":[67],"processes":[69],"MHSA":[70],"using":[72,114],"cascaded":[73],"hardware":[74],"modules:":[75],"Query-Key-Value":[78],"generator":[79],"(SpikeQKV),":[80],"low-complexity":[82],"Multi-Head":[84],"Self-Attention":[85],"engine":[86],"(SpikeAtten),":[87],"Reparameterization":[89],"Convolution":[90],"(RepConv)":[91],"modules.":[92],"mitigate":[94],"traffic":[95],"congestion":[96],"on-chip":[98],"memory":[99],"accesses":[100],"data":[102],"reuse,":[103],"specialized":[104],"inter-module":[105],"buffers":[106],"integrated":[108],"within":[109],"dataflow.":[111],"Under":[112],"synthesis":[113],"32nm":[115],"CMOS":[116],"technology,":[117],"MorphAtt":[118,150],"achieves":[119],"792-1605":[120],"GOPS":[121],"throughput,":[123],"while":[124],"incurring":[125],"~39-55":[126],"mW":[127],"power":[129],"consumption":[130],"1.5":[132],"mm^2":[133],"area,":[135],"which":[136],"lead":[137],"20.3-29.1":[139],"TOPS/W":[140],"efficiency.":[143],"These":[144],"results":[145],"also":[146],"demonstrate":[147],"our":[149],"offers":[151],"better":[152],"performance":[153],"trade-offs":[156],"than":[157],"state-of-the-art,":[158],"thereby":[159],"enabling":[160],"highly":[161],"vision-based":[163],"AI":[164],"systems":[165]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-09-30T00:00:00"}