{"id":"https://openalex.org/W7223835010","doi":"https://doi.org/10.5281/zenodo.23689194","title":"OTAZ-DualPath: A Proprietary Inference Engine for Sparse Quantized Neural Networks on Low-Power CPUs","display_name":"OTAZ-DualPath: A Proprietary Inference Engine for Sparse Quantized Neural Networks on Low-Power CPUs","publication_year":2026,"publication_date":"2026-10-26","ids":{"openalex":"https://openalex.org/W7223835010","doi":"https://doi.org/10.5281/zenodo.23689194"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23689194","doi":"https://doi.org/10.5281/zenodo.23689194","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23689194","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23689194","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5159874270","display_name":"Al-Dawoodiya Al-Mu'taz Billah Salah","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Al-Mu'taz Billah Salah, Al-Dawoodiya","raw_affiliation_strings":["Independent Researcher / OTAZ-DualPath, Amman, Jordan","Independent Researcher","OTAZ-DualPath, Amman, Jordan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent Researcher / OTAZ-DualPath, Amman, Jordan","institution_ids":[]},{"raw_affiliation_string":"Independent Researcher","institution_ids":[]},{"raw_affiliation_string":"OTAZ-DualPath, Amman, Jordan","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5159874270"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9663000106811523,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9663000106811523,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.019099999219179153,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.00570000009611249,"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/weight-pruning","display_name":"weight pruning","score":0.6830000281333923},{"id":"https://openalex.org/keywords/post-training-quantization","display_name":"post-training quantization","score":0.8389999866485596},{"id":"https://openalex.org/keywords/memory-compression","display_name":"memory compression","score":0.847000002861023},{"id":"https://openalex.org/keywords/edge-intelligence","display_name":"edge AI","score":0.7590000033378601},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.6639999747276306}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8087999820709229},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5449000000953674},{"id":"https://openalex.org/C150552126","wikidata":"https://www.wikidata.org/wiki/Q339387","display_name":"SIMD","level":2,"score":0.5382999777793884},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4587000012397766},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.45179998874664307},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.4438999891281128},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.43650001287460327},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4153999984264374},{"id":"https://openalex.org/C170723468","wikidata":"https://www.wikidata.org/wiki/Q182933","display_name":"x86","level":3,"score":0.41130000352859497},{"id":"https://openalex.org/C46743427","wikidata":"https://www.wikidata.org/wiki/Q1341685","display_name":"Inference engine","level":3,"score":0.39899998903274536},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.39100000262260437},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3817000091075897},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.3506999909877777},{"id":"https://openalex.org/C2778319340","wikidata":"https://www.wikidata.org/wiki/Q776861","display_name":"Transputer","level":2,"score":0.33160001039505005},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.3271999955177307},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.3224000036716461},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.32190001010894775},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.31060001254081726},{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.30379998683929443},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.296999990940094},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.2874000072479248},{"id":"https://openalex.org/C53442348","wikidata":"https://www.wikidata.org/wiki/Q745101","display_name":"Content-addressable memory","level":3,"score":0.28529998660087585},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.27869999408721924},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.2720000147819519},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C107027933","wikidata":"https://www.wikidata.org/wiki/Q2006448","display_name":"Stream processing","level":2,"score":0.26759999990463257},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.25929999351501465},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.2540999948978424},{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.25290000438690186},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.25220000743865967},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.2515000104904175}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23689194","doi":"https://doi.org/10.5281/zenodo.23689194","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23689194","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.23689196","doi":"https://doi.org/10.5281/zenodo.23689196","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23689196","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"doi:10.5281/zenodo.23689194","doi":"https://doi.org/10.5281/zenodo.23689194","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23689194","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4412397574","https://openalex.org/W7128789769","https://openalex.org/W4393147284","https://openalex.org/W7172407000","https://openalex.org/W7169586281","https://openalex.org/W3125082172","https://openalex.org/W7130529450","https://openalex.org/W7203694013","https://openalex.org/W4416247035","https://openalex.org/W7151317453"],"abstract_inverted_index":{"Running":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"and":[5,24,66,116,142,171,190,229,249,261,266],"deep":[6],"neural":[7],"networks":[8],"on":[9,69,162,204,224],"low-power":[10,84],"central":[11],"processing":[12],"units":[13],"(CPUs)":[14],"remains":[15],"a":[16,76,92,128,234],"critical":[17,99],"challenge":[18],"for":[19,43,83,181,239],"edge":[20,250],"AI,":[21],"embedded":[22,246],"systems,":[23],"cost-sensitive":[25],"AI":[26,242],"infrastructure.":[27],"Traditional":[28],"quantization":[29,193,265],"methods":[30],"apply":[31],"uniform":[32],"precision":[33],"reduction":[34],"to":[35,40,110,121,158,173,241],"all":[36],"model":[37,154,217,227],"weights,":[38],"leading":[39],"accuracy":[41,155],"degradation\u2014particularly":[42],"outlier":[44,100],"weights":[45,101,109,120],"that":[46,149],"carry":[47],"disproportionate":[48],"importance.":[49],"Furthermore,":[50],"existing":[51,264],"solutions":[52],"often":[53],"require":[54],"Graphics":[55],"Processing":[56],"Units":[57],"(GPUs),":[58],"calibration":[59,188],"datasets,":[60],"or":[61],"suffer":[62],"from":[63],"high":[64],"memory":[65,160,169],"compute":[67],"overhead":[68,141],"CPU":[70,209],"hardware.":[71],"This":[72,253],"paper":[73],"presents":[74],"OTAZ-DualPath,":[75],"proprietary":[77,235],"inference":[78,267],"execution":[79,130],"engine":[80,185],"designed":[81],"specifically":[82],"x86":[85],"CPUs":[86],"(15W\u201365W":[87],"TDP":[88],"class).":[89],"OTAZ-DualPath":[90,150,232],"introduces":[91],"three-tier":[93],"adaptive":[94,113],"weight":[95],"classification:":[96],"Tier-1":[97],"preserves":[98],"in":[102,194],"full":[103],"FP32":[104],"precision,":[105],"Tier-2":[106],"quantizes":[107],"normal":[108],"Int8":[111],"using":[112],"per-row":[114],"scaling,":[115],"Tier-3":[117],"prunes":[118],"unimportant":[119],"zero.":[122],"All":[123,197],"values":[124],"are":[125,200],"unified":[126],"into":[127],"single":[129],"stream":[131],"processed":[132],"by":[133],"one":[134],"optimized":[135],"SIMD":[136],"kernel":[137],"(AVX2+FMA),":[138],"eliminating":[139],"branching":[140],"dual-path":[143],"execution.":[144],"Illustrative":[145],"theoretical":[146,201],"results":[147,220],"demonstrate":[148],"can":[151],"achieve":[152],"near-FP32":[153],"with":[156,210,263],"up":[157,172],"7.7\u00d7":[159],"compression":[161],"sparse":[163],"models":[164],"(90%":[165],"sparsity),":[166],"zero":[167],"runtime":[168],"allocations,":[170],"5.5\u00d7":[174],"parallelism":[175],"efficiency":[176],"improvement":[177],"over":[178],"standard":[179],"OpenMP":[180],"small-batch":[182],"inference.":[183],"The":[184],"requires":[186],"no":[187],"data":[189],"performs":[191],"post-training":[192],"seconds.":[195],"Note:":[196],"numerical":[198],"examples":[199],"estimates":[202],"based":[203],"specific":[205,216],"experimental":[206],"conditions":[207],"(x86":[208],"AVX2+FMA,":[211],"batch":[212],"size":[213],"=":[214],"1,":[215],"dimensions).":[218],"Actual":[219],"may":[221],"vary":[222],"depending":[223],"hardware":[225],"configuration,":[226],"architecture,":[228,257],"deployment":[230],"scenario.":[231],"is":[233],"commercial":[236],"product":[237],"available":[238],"licensing":[240],"companies,":[243],"IoT":[244],"manufacturers,":[245],"device":[247],"OEMs,":[248],"computing":[251],"providers.":[252],"preprint":[254],"outlines":[255],"the":[256],"features,":[258],"performance":[259],"results,":[260],"comparison":[262],"systems.":[268]},"counts_by_year":[],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2026-10-28T00:00:00"}