{"id":"https://openalex.org/W7210284750","doi":"https://doi.org/10.58445/rars.4138","title":"Engineering AI Accelerators: GPUs, TPUs, and NPUs","display_name":"Engineering AI Accelerators: GPUs, TPUs, and NPUs","publication_year":2026,"publication_date":"2026-09-06","ids":{"openalex":"https://openalex.org/W7210284750","doi":"https://doi.org/10.58445/rars.4138"},"language":"en","primary_location":{"id":"doi:10.58445/rars.4138","doi":"https://doi.org/10.58445/rars.4138","is_oa":true,"landing_page_url":"https://doi.org/10.58445/rars.4138","pdf_url":"https://research-archive.org/index.php/rars/preprint/download/4138/5676/5250","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://research-archive.org/index.php/rars/preprint/download/4138/5676/5250","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5146571604","display_name":"Yongjian Zhang","orcid":"https://orcid.org/0000-0002-6178-6532","observed_orcids":["https://orcid.org/0000-0002-6178-6532"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yongjian Zhang","raw_affiliation_strings":["student"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"student","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5146571604"],"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9682999849319458,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9682999849319458,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.017899999395012856,"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/T13904","display_name":"Artificial Intelligence Applications","score":0.004100000020116568,"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/ai-hardware","display_name":"AI accelerators","score":0.9940000176429749},{"id":"https://openalex.org/keywords/graphics-processing-units","display_name":"graphics processing units","score":0.8069999814033508},{"id":"https://openalex.org/keywords/tensor-processing-units","display_name":"tensor processing units","score":1.0},{"id":"https://openalex.org/keywords/neural-processing-units","display_name":"neural processing units","score":0.9990000128746033},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"parallel computing","score":0.7450000047683716},{"id":"https://openalex.org/keywords/machine-learning-hardware","display_name":"machine learning hardware","score":0.8009999990463257},{"id":"https://openalex.org/keywords/matrix-operations","display_name":"matrix operations","score":0.8799999952316284},{"id":"https://openalex.org/keywords/mobile-devices","display_name":"mobile devices","score":0.42100000381469727}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7035999894142151},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5297999978065491},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.503600001335144},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.47119998931884766},{"id":"https://openalex.org/C157170001","wikidata":"https://www.wikidata.org/wiki/Q4781507","display_name":"Applications of artificial intelligence","level":2,"score":0.44670000672340393},{"id":"https://openalex.org/C138827492","wikidata":"https://www.wikidata.org/wiki/Q6661985","display_name":"Data processing","level":2,"score":0.42559999227523804},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.38280001282691956},{"id":"https://openalex.org/C106515295","wikidata":"https://www.wikidata.org/wiki/Q26806595","display_name":"Parallel processing","level":2,"score":0.38029998540878296},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.3393000066280365},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.32420000433921814},{"id":"https://openalex.org/C87868495","wikidata":"https://www.wikidata.org/wiki/Q750843","display_name":"Information processing","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C86111242","wikidata":"https://www.wikidata.org/wiki/Q859595","display_name":"Coprocessor","level":2,"score":0.2921999990940094},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.2897000014781952},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.2782000005245209},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2655999958515167},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.2612000107765198},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.25619998574256897}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.58445/rars.4138","doi":"https://doi.org/10.58445/rars.4138","is_oa":true,"landing_page_url":"https://doi.org/10.58445/rars.4138","pdf_url":"https://research-archive.org/index.php/rars/preprint/download/4138/5676/5250","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"}],"best_oa_location":{"id":"doi:10.58445/rars.4138","doi":"https://doi.org/10.58445/rars.4138","is_oa":true,"landing_page_url":"https://doi.org/10.58445/rars.4138","pdf_url":"https://research-archive.org/index.php/rars/preprint/download/4138/5676/5250","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7210284750.pdf","grobid_xml":"https://content.openalex.org/works/W7210284750.grobid-xml"},"referenced_works_count":1,"referenced_works":["https://openalex.org/W4404954356"],"related_works":["https://openalex.org/W4417495754","https://openalex.org/W4393104983","https://openalex.org/W4396686371","https://openalex.org/W7219792329","https://openalex.org/W4406878399","https://openalex.org/W3158985940","https://openalex.org/W4411535957","https://openalex.org/W4327511194","https://openalex.org/W7217416156","https://openalex.org/W3032230397"],"abstract_inverted_index":{"Artificial":[0],"intelligence":[1,107],"requires":[2],"specialized":[3,61,102],"hardware":[4,104],"to":[5,42,53,108],"perform":[6],"the":[7,85],"large":[8],"number":[9],"of":[10,23,89],"mathematical":[11],"calculations":[12],"needed":[13],"for":[14,62,71],"training":[15],"and":[16,34,58,66,76,87,92,111],"inference.This":[17],"paper":[18,79],"examines":[19],"three":[20],"major":[21],"types":[22],"AI":[24,44,100,103],"accelerators:":[25],"Graphics":[26],"Processing":[27,31,36],"Units":[28,32,37],"(GPUs),":[29],"Tensor":[30],"(TPUs),":[33],"Neural":[35],"(NPUs).Each":[38],"processor":[39],"is":[40],"designed":[41,70],"improve":[43],"performance":[45,57],"in":[46,98],"different":[47],"ways.GPUs":[48],"are":[49,60,69],"widely":[50],"used":[51],"due":[52],"their":[54],"parallel":[55],"processing":[56,73],"flexibility.TPUs":[59],"machine":[63],"learning":[64],"workloads":[65],"matrix":[67],"operations.NPUs":[68],"efficient":[72],"on":[74,117],"laptops":[75],"cell":[77],"phones.This":[78],"explains":[80],"how":[81,93],"each":[82,90,94],"accelerator":[83],"works,":[84],"advantages":[86],"disadvantages":[88],"accelerator,":[91],"plays":[95],"a":[96],"role":[97],"modern":[99],"systems.Overall,":[101],"allows":[105],"artificial":[106],"operate":[109],"faster":[110],"more":[112],"efficiently":[113],"compared":[114],"with":[115],"relying":[116],"CPUs":[118],"alone.":[119]},"counts_by_year":[],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2026-09-07T00:00:00"}