{"id":"https://openalex.org/W4416004231","doi":"https://doi.org/10.1145/3731599.3767395","title":"Enabling Unstructured Sparse Fine-Tuning and Inference for Foundation Models on Wafer-Scale Engine","display_name":"Enabling Unstructured Sparse Fine-Tuning and Inference for Foundation Models on Wafer-Scale Engine","publication_year":2025,"publication_date":"2025-11-07","ids":{"openalex":"https://openalex.org/W4416004231","doi":"https://doi.org/10.1145/3731599.3767395"},"language":null,"primary_location":{"id":"doi:10.1145/3731599.3767395","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767395","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3731599.3767395","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5075144202","display_name":"Haoyu Zheng","orcid":"https://orcid.org/0000-0003-1319-6439"},"institutions":[{"id":"https://openalex.org/I131249849","display_name":"Oregon State University","ror":"https://ror.org/00ysfqy60","country_code":"US","type":"education","lineage":["https://openalex.org/I131249849"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Haoyu Zheng","raw_affiliation_strings":["Oregon State University, Corvallis, USA"],"raw_orcid":"https://orcid.org/0000-0003-1319-6439","affiliations":[{"raw_affiliation_string":"Oregon State University, Corvallis, USA","institution_ids":["https://openalex.org/I131249849"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034872239","display_name":"Yifan Zeng","orcid":"https://orcid.org/0000-0002-3174-9255"},"institutions":[{"id":"https://openalex.org/I131249849","display_name":"Oregon State University","ror":"https://ror.org/00ysfqy60","country_code":"US","type":"education","lineage":["https://openalex.org/I131249849"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yifan Zeng","raw_affiliation_strings":["Oregon State University, Corvallis, USA"],"raw_orcid":"https://orcid.org/0000-0002-3174-9255","affiliations":[{"raw_affiliation_string":"Oregon State University, Corvallis, USA","institution_ids":["https://openalex.org/I131249849"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036239104","display_name":"Linghao Song","orcid":"https://orcid.org/0000-0002-7450-2842"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Linghao Song","raw_affiliation_strings":["Yale University, New Haven, USA"],"raw_orcid":"https://orcid.org/0000-0002-7450-2842","affiliations":[{"raw_affiliation_string":"Yale University, New Haven, USA","institution_ids":["https://openalex.org/I32971472"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014448094","display_name":"Murali Emani","orcid":"https://orcid.org/0000-0002-6279-0007"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Murali Emani","raw_affiliation_strings":["Argonne National Laboratory, Chicago, USA"],"raw_orcid":"https://orcid.org/0000-0002-6279-0007","affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, Chicago, USA","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102994344","display_name":"Wenqian Dong","orcid":"https://orcid.org/0000-0002-8376-6647"},"institutions":[{"id":"https://openalex.org/I131249849","display_name":"Oregon State University","ror":"https://ror.org/00ysfqy60","country_code":"US","type":"education","lineage":["https://openalex.org/I131249849"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wenqian Dong","raw_affiliation_strings":["Oregon State University, Corvallis, USA"],"raw_orcid":"https://orcid.org/0000-0002-8376-6647","affiliations":[{"raw_affiliation_string":"Oregon State University, Corvallis, USA","institution_ids":["https://openalex.org/I131249849"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5075144202"],"corresponding_institution_ids":["https://openalex.org/I131249849"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.24244789,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"500","last_page":"507"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.45410001277923584,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.45410001277923584,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.09790000319480896,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.09309999644756317,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.8794000148773193},{"id":"https://openalex.org/keywords/inference-engine","display_name":"Inference engine","score":0.7423999905586243},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5149000287055969},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.45829999446868896},{"id":"https://openalex.org/keywords/phoenix","display_name":"Phoenix","score":0.45730000734329224},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.4431999921798706},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.4047999978065491},{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.38589999079704285},{"id":"https://openalex.org/keywords/downstream","display_name":"Downstream (manufacturing)","score":0.33880001306533813}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.8794000148773193},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7878999710083008},{"id":"https://openalex.org/C46743427","wikidata":"https://www.wikidata.org/wiki/Q1341685","display_name":"Inference engine","level":3,"score":0.7423999905586243},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5149000287055969},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.45829999446868896},{"id":"https://openalex.org/C2779167034","wikidata":"https://www.wikidata.org/wiki/Q27685","display_name":"Phoenix","level":3,"score":0.45730000734329224},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4496999979019165},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.44769999384880066},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4431999921798706},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.4047999978065491},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.38589999079704285},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.33880001306533813},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.3352000117301941},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3255000114440918},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.3246999979019165},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.304500013589859},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.30079999566078186},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.29330000281333923},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.28839999437332153},{"id":"https://openalex.org/C2781252014","wikidata":"https://www.wikidata.org/wiki/Q1141900","display_name":"Unstructured data","level":3,"score":0.27869999408721924},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.2759999930858612},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.2745000123977661},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.27390000224113464},{"id":"https://openalex.org/C198370458","wikidata":"https://www.wikidata.org/wiki/Q586459","display_name":"Type inference","level":3,"score":0.267300009727478},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.25760000944137573},{"id":"https://openalex.org/C207685749","wikidata":"https://www.wikidata.org/wiki/Q2088941","display_name":"Domain knowledge","level":2,"score":0.25290000438690186}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3731599.3767395","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767395","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3731599.3767395","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767395","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":13,"referenced_works":["https://openalex.org/W2475046758","https://openalex.org/W2890894339","https://openalex.org/W2946609015","https://openalex.org/W2946659172","https://openalex.org/W3170403598","https://openalex.org/W3194676777","https://openalex.org/W4297097426","https://openalex.org/W4376130831","https://openalex.org/W4401042761","https://openalex.org/W4401861759","https://openalex.org/W4402670757","https://openalex.org/W4404781719","https://openalex.org/W6893007900"],"related_works":[],"abstract_inverted_index":{"Foundation":[0],"models,":[1,6,58,130],"such":[2],"as":[3],"large":[4],"language":[5],"have":[7],"become":[8],"ubiquitous":[9],"and":[10,131,166],"are":[11,18],"employed":[12],"in":[13,56,66,164,171],"many":[14],"applications.":[15],"These":[16],"models":[17],"often":[19],"adapted":[20],"to":[21,60,75,96,116,137,147,158,175],"a":[22,29,38,73,132,159,167,176],"desired":[23],"domain":[24],"or":[25],"downstream":[26,80],"task":[27],"through":[28],"fine-tuning":[30,54],"stage.":[31,184],"Cerebras,":[32],"the":[33,62,103,111,118,139,182],"wafer-scale":[34,114],"AI":[35],"accelerator,":[36],"is":[37],"suitable":[39],"platform":[40],"for":[41,79],"running":[42],"foundation":[43],"models.":[44],"However,":[45],"existing":[46],"works":[47],"utilizing":[48],"sparse":[49,77,91],"pre-training":[50],"followed":[51],"by":[52],"dense":[53,57,148],"result":[55],"failing":[59],"reduce":[61],"inference":[63,78,119,172,183],"cost,":[64],"measured":[65],"floating-point":[67],"operations.":[68],"This":[69],"paper":[70],"proposes":[71],"Phoenix,":[72],"solution":[74],"enable":[76,97],"tasks":[81],"on":[82],"Cerebras.":[83],"Phoenix":[84,108,143],"features":[85],"an":[86],"innovative":[87],"strategy":[88],"that":[89],"merges":[90],"weights":[92],"with":[93],"low-rank":[94],"adapters":[95],"sparsity":[98,123,135],"without":[99],"losing":[100],"accuracy,":[101],"overcoming":[102],"limitations":[104],"of":[105,134,141],"previous":[106],"approaches.":[107],"also":[109],"leverages":[110],"Cerebras":[112],"CS-2":[113],"engine":[115],"accelerate":[117],"process":[120],"using":[121],"unstructured":[122],"methods.":[124],"We":[125],"evaluate":[126],"multiple":[127],"adaptation":[128],"scenarios,":[129],"range":[133],"levels":[136],"demonstrate":[138],"effectiveness":[140],"Phoenix.":[142],"maintains":[144],"accuracy":[145],"comparable":[146],"baselines":[149],"even":[150],"at":[151],"50\u201360%":[152],"sparsity.":[153],"Meanwhile,":[154],"it":[155],"achieves":[156],"up":[157],"near":[160],"2":[161],"\u00d7":[162,169],"reduction":[163],"FLOPs":[165],"1.7":[168],"improvement":[170],"throughput":[173],"compared":[174],"single":[177],"NVIDIA":[178],"A100":[179],"GPU":[180],"during":[181]},"counts_by_year":[],"updated_date":"2025-11-08T23:21:52.890332","created_date":"2025-11-07T00:00:00"}