{"id":"https://openalex.org/W4408029577","doi":"https://doi.org/10.1145/3710848.3710871","title":"MARLIN: Mixed-Precision Auto-Regressive Parallel Inference on Large Language Models","display_name":"MARLIN: Mixed-Precision Auto-Regressive Parallel Inference on Large Language Models","publication_year":2025,"publication_date":"2025-02-28","ids":{"openalex":"https://openalex.org/W4408029577","doi":"https://doi.org/10.1145/3710848.3710871"},"language":"en","primary_location":{"id":"doi:10.1145/3710848.3710871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3710848.3710871","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3710848.3710871","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 30th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3710848.3710871","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5060660774","display_name":"Elias Frantar","orcid":null},"institutions":[{"id":"https://openalex.org/I157556583","display_name":"Institute of Science and Technology Austria","ror":"https://ror.org/03gnh5541","country_code":"AT","type":"education","lineage":["https://openalex.org/I157556583"]}],"countries":["AT"],"is_corresponding":true,"raw_author_name":"Elias Frantar","raw_affiliation_strings":["ISTA, Klosterneuburg, Austria"],"affiliations":[{"raw_affiliation_string":"ISTA, Klosterneuburg, Austria","institution_ids":["https://openalex.org/I157556583"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005191985","display_name":"Roberto L. Castro","orcid":"https://orcid.org/0000-0001-5493-0287"},"institutions":[{"id":"https://openalex.org/I11019714","display_name":"Universidade da Coru\u00f1a","ror":"https://ror.org/01qckj285","country_code":"ES","type":"education","lineage":["https://openalex.org/I11019714"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Roberto L. Castro","raw_affiliation_strings":["CITIC, Universidade da Coru\u00f1a, A Coru\u00f1a, Spain"],"affiliations":[{"raw_affiliation_string":"CITIC, Universidade da Coru\u00f1a, A Coru\u00f1a, Spain","institution_ids":["https://openalex.org/I11019714"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100696894","display_name":"Jiale Chen","orcid":"https://orcid.org/0000-0001-5337-5875"},"institutions":[{"id":"https://openalex.org/I157556583","display_name":"Institute of Science and Technology Austria","ror":"https://ror.org/03gnh5541","country_code":"AT","type":"education","lineage":["https://openalex.org/I157556583"]}],"countries":["AT"],"is_corresponding":false,"raw_author_name":"Jiale Chen","raw_affiliation_strings":["ISTA, Klosterneuburg, Austria"],"affiliations":[{"raw_affiliation_string":"ISTA, Klosterneuburg, Austria","institution_ids":["https://openalex.org/I157556583"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026990786","display_name":"Torsten Hoefler","orcid":"https://orcid.org/0000-0002-1333-9797"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Torsten Hoefler","raw_affiliation_strings":["ETH Z\u00fcrich, Z\u00fcrich, Switzerland"],"affiliations":[{"raw_affiliation_string":"ETH Z\u00fcrich, Z\u00fcrich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5083822059","display_name":"Dan Alistarh","orcid":"https://orcid.org/0000-0003-3650-940X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dan Alistarh","raw_affiliation_strings":["ISTA, Klosterneuburg, Austria, Neural Magic, Inc., Somerville, USA"],"affiliations":[{"raw_affiliation_string":"ISTA, Klosterneuburg, Austria, Neural Magic, Inc., Somerville, USA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5060660774"],"corresponding_institution_ids":["https://openalex.org/I157556583"],"apc_list":null,"apc_paid":null,"fwci":32.3034,"has_fulltext":false,"cited_by_count":13,"citation_normalized_percentile":{"value":0.99581288,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"239","last_page":"251"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9986000061035156,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9986000061035156,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9902999997138977,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T13702","display_name":"Machine Learning in Healthcare","score":0.9868000149726868,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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":0.6656580567359924},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6541733741760254},{"id":"https://openalex.org/keywords/autoregressive-model","display_name":"Autoregressive model","score":0.5910189747810364},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.30075281858444214},{"id":"https://openalex.org/keywords/econometrics","display_name":"Econometrics","score":0.22914502024650574},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.12646570801734924}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6656580567359924},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6541733741760254},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.5910189747810364},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.30075281858444214},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.22914502024650574},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.12646570801734924}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3710848.3710871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3710848.3710871","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3710848.3710871","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 30th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming","raw_type":"proceedings-article"},{"id":"pmh:doi:10.3929/ethz-b-000729444","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Conference Paper"}],"best_oa_location":{"id":"doi:10.1145/3710848.3710871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3710848.3710871","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3710848.3710871","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 30th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4408029577.pdf"},"referenced_works_count":3,"referenced_works":["https://openalex.org/W4309672181","https://openalex.org/W4321446237","https://openalex.org/W4387321091"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2171218219","https://openalex.org/W1972271943","https://openalex.org/W2150410159","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345"],"abstract_inverted_index":{"As":[0],"inference":[1],"on":[2],"Large":[3],"Language":[4],"Models":[5],"(LLMs)":[6],"emerges":[7],"as":[8],"an":[9],"important":[10],"workload":[11],"in":[12,67],"machine":[13],"learning":[14],"applications,":[15],"model":[16,32],"weight":[17],"quantization":[18],"has":[19,35],"become":[20],"a":[21,58],"standard":[22],"technique":[23],"for":[24,43,78],"efficient":[25],"GPU":[26,85],"deployment.":[27],"Quantization":[28],"not":[29],"only":[30],"reduces":[31],"size,":[33],"but":[34],"also":[36,66],"been":[37],"shown":[38],"to":[39,47,90],"yield":[40],"substantial":[41],"speedups":[42,63],"single-user":[44],"inference,":[45],"due":[46],"reduced":[48],"memory":[49],"movement,":[50],"with":[51,70],"low":[52],"accuracy":[53],"impact.":[54],"Yet,":[55],"it":[56],"remains":[57],"key":[59],"open":[60],"question":[61],"whether":[62,84],"are":[64,75],"achievable":[65],"batched":[68,102],"settings":[69],"multiple":[71],"parallel":[72],"clients,":[73],"which":[74],"highly":[76],"relevant":[77],"practical":[79],"serving.":[80],"It":[81],"is":[82],"unclear":[83],"kernels":[86],"can":[87],"be":[88],"designed":[89],"remain":[91],"practically":[92],"memory-bound,":[93],"while":[94],"supporting":[95],"the":[96],"substantially":[97],"increased":[98],"compute":[99],"requirements":[100],"of":[101],"workloads.":[103]},"counts_by_year":[{"year":2026,"cited_by_count":8},{"year":2025,"cited_by_count":5}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}