{"id":"https://openalex.org/W7212027134","doi":"https://doi.org/10.48550/arxiv.2609.05709","title":"cuSOAP: a GPU-accelerated Generator of Smooth Overlap of Atomic Positions Descriptor","display_name":"cuSOAP: a GPU-accelerated Generator of Smooth Overlap of Atomic Positions Descriptor","publication_year":2026,"publication_date":"2026-09-04","ids":{"openalex":"https://openalex.org/W7212027134","doi":"https://doi.org/10.48550/arxiv.2609.05709"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.05709","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.05709","pdf_url":"https://arxiv.org/pdf/2609.05709","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.05709","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5076132813","display_name":"Hongyu Yan","orcid":"https://orcid.org/0009-0002-0454-2256","observed_orcids":["https://orcid.org/0009-0002-0454-2256"]},"institutions":[{"id":"https://openalex.org/I32062511","display_name":"Heriot-Watt University","ror":"https://ror.org/04mghma93","country_code":"GB","type":"education","lineage":["https://openalex.org/I32062511"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Yan, Hongyu","raw_affiliation_strings":["School of Engineering and Physical Sciences , Heriot-Watt University , Edinburgh EH14 4AS , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Engineering and Physical Sciences , Heriot-Watt University , Edinburgh EH14 4AS , UK","institution_ids":["https://openalex.org/I32062511"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5150940615","display_name":"Yuqing Xia","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Xia, Yuqing","raw_affiliation_strings":["School of Artificial Intelligence , Chinese University of Hong Kong , Shatin , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Artificial Intelligence , Chinese University of Hong Kong , Shatin , Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5150890886","display_name":"Yong Wei","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I924760719","display_name":"University of North Georgia","ror":"https://ror.org/001pe5g24","country_code":"US","type":"education","lineage":["https://openalex.org/I924760719"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wei, Yong","raw_affiliation_strings":["Department of Computer Science and Cybersecurity , University of North Georgia , Dahlonega , Georgia 30597 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Cybersecurity , University of North Georgia , Dahlonega , Georgia 30597 , USA","institution_ids":["https://openalex.org/I924760719"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5150889777","display_name":"Minghan Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47251452","display_name":"Wake Forest University","ror":"https://ror.org/0207ad724","country_code":"US","type":"education","lineage":["https://openalex.org/I47251452"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chen, Minghan","raw_affiliation_strings":["Department of Computer Science , Wake Forest University , Winston-Salem , North Carolina 27104 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Wake Forest University , Winston-Salem , North Carolina 27104 , USA","institution_ids":["https://openalex.org/I47251452"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5151070207","display_name":"Hanning Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4388891828","display_name":"Texas Advanced Computing Center","ror":"https://ror.org/00xg4bh43","country_code":"US","type":"facility","lineage":["https://openalex.org/I4388891828","https://openalex.org/I86519309"]},{"id":"https://openalex.org/I86519309","display_name":"The University of Texas at Austin","ror":"https://ror.org/00hj54h04","country_code":"US","type":"education","lineage":["https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chen, Hanning","raw_affiliation_strings":["Texas Advanced Computing Center , The University of Texas at Austin , Austin , Texas 78758 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Texas Advanced Computing Center , The University of Texas at Austin , Austin , Texas 78758 , USA","institution_ids":["https://openalex.org/I4388891828","https://openalex.org/I86519309"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":6,"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/T11948","display_name":"Machine Learning in Materials Science","score":0.9398999810218811,"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":0.9398999810218811,"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/T10002","display_name":"Advanced Chemical Physics Studies","score":0.0430000014603138,"subfield":{"id":"https://openalex.org/subfields/3107","display_name":"Atomic and Molecular Physics, and Optics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.011500000022351742,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/soap-descriptors","display_name":"SOAP descriptors","score":0.8659999966621399},{"id":"https://openalex.org/keywords/gpu-acceleration","display_name":"GPU acceleration","score":0.6129999756813049},{"id":"https://openalex.org/keywords/interatomic-potentials","display_name":"interatomic potentials","score":0.8569999933242798},{"id":"https://openalex.org/keywords/molecular-machine-learning","display_name":"molecular machine learning","score":0.718999981880188},{"id":"https://openalex.org/keywords/analytic-derivatives","display_name":"analytical derivatives","score":0.44699999690055847},{"id":"https://openalex.org/keywords/gaussian-type-orbitals","display_name":"Gaussian-type orbitals","score":0.7889999747276306},{"id":"https://openalex.org/keywords/triton-kernels","display_name":"Triton kernels","score":0.6959999799728394},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.531000018119812},{"id":"https://openalex.org/keywords/water-clusters","display_name":"water clusters","score":0.6370000243186951},{"id":"https://openalex.org/keywords/condensed-phase-simulation","display_name":"condensed phase simulation","score":0.7289999723434448},{"id":"https://openalex.org/keywords/parallel-efficiency","display_name":"parallel efficiency","score":0.4830000102519989}],"concepts":[{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6195999979972839},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.585099995136261},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.5503000020980835},{"id":"https://openalex.org/C2780992000","wikidata":"https://www.wikidata.org/wiki/Q17016113","display_name":"Generator (circuit theory)","level":3,"score":0.5429999828338623},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.4814000129699707},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4462999999523163},{"id":"https://openalex.org/C42747912","wikidata":"https://www.wikidata.org/wiki/Q1048447","display_name":"Multiplicative function","level":2,"score":0.4088999927043915},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4059999883174896},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.4058000147342682},{"id":"https://openalex.org/C12426560","wikidata":"https://www.wikidata.org/wiki/Q189569","display_name":"Basis (linear algebra)","level":2,"score":0.39410001039505005},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.37720000743865967},{"id":"https://openalex.org/C133095886","wikidata":"https://www.wikidata.org/wiki/Q1307173","display_name":"Single-precision floating-point format","level":3,"score":0.37689998745918274},{"id":"https://openalex.org/C139676723","wikidata":"https://www.wikidata.org/wiki/Q1193832","display_name":"Sign (mathematics)","level":2,"score":0.36469998955726624},{"id":"https://openalex.org/C86111242","wikidata":"https://www.wikidata.org/wiki/Q859595","display_name":"Coprocessor","level":2,"score":0.34610000252723694},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.34529998898506165},{"id":"https://openalex.org/C5917680","wikidata":"https://www.wikidata.org/wiki/Q2621825","display_name":"Basis function","level":2,"score":0.3319000005722046},{"id":"https://openalex.org/C32900221","wikidata":"https://www.wikidata.org/wiki/Q181365","display_name":"Dot product","level":2,"score":0.32749998569488525},{"id":"https://openalex.org/C16038011","wikidata":"https://www.wikidata.org/wiki/Q62912","display_name":"Cartesian coordinate system","level":2,"score":0.3084000051021576},{"id":"https://openalex.org/C2778770139","wikidata":"https://www.wikidata.org/wiki/Q1966904","display_name":"Solver","level":2,"score":0.30329999327659607},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.2915000021457672},{"id":"https://openalex.org/C35912277","wikidata":"https://www.wikidata.org/wiki/Q1243369","display_name":"Double-precision floating-point format","level":3,"score":0.2842999994754791},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.26919999718666077},{"id":"https://openalex.org/C30475298","wikidata":"https://www.wikidata.org/wiki/Q909554","display_name":"Computational physics","level":1,"score":0.26179999113082886},{"id":"https://openalex.org/C90119067","wikidata":"https://www.wikidata.org/wiki/Q43260","display_name":"Polynomial","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C81063470","wikidata":"https://www.wikidata.org/wiki/Q1256787","display_name":"Rotational speed","level":2,"score":0.2581000030040741}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.05709","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.05709","pdf_url":"https://arxiv.org/pdf/2609.05709","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.2609.05709","doi":"https://doi.org/10.48550/arxiv.2609.05709","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.05709","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.05709","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.05709","pdf_url":"https://arxiv.org/pdf/2609.05709","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":[{"id":"https://openalex.org/F4320310620","display_name":"University of Texas at Austin","ror":"https://ror.org/00hj54h04"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7212027134.pdf","grobid_xml":"https://content.openalex.org/works/W7212027134.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2943241318","https://openalex.org/W4365144343","https://openalex.org/W2955413888","https://openalex.org/W2346508658","https://openalex.org/W4388412908","https://openalex.org/W4412167617","https://openalex.org/W4290805632","https://openalex.org/W4321793397","https://openalex.org/W4381309085","https://openalex.org/W4385725155"],"abstract_inverted_index":{"The":[0,108],"Smooth":[1],"Overlap":[2],"of":[3,10,16,26,36,61,103,165,228,238],"Atomic":[4],"Positions":[5],"(SOAP)":[6],"descriptor":[7],"is":[8,110,209],"one":[9,214],"the":[11,24,42,79,100,115,152,176,183,187,225,229],"most":[12],"widely":[13],"adopted":[14],"representations":[15],"atomic":[17],"environments":[18],"in":[19,211,222],"molecular":[20],"machine":[21],"learning,":[22],"but":[23],"cost":[25],"evaluating":[27],"it":[28],"and":[29,45,65,82,88,95,125,130],"its":[30,120],"derivatives":[31],"remains":[32],"a":[33,58,104,111,147,157,204,218],"principal":[34],"bottleneck":[35],"SOAP-based":[37],"interatomic":[38],"potentials,":[39],"particularly":[40],"at":[41],"large":[43],"radial":[44,90],"angular":[46,188],"basis":[47],"sizes":[48],"demanded":[49],"by":[50],"complex,":[51],"multi-species":[52],"condensed-phase":[53,252],"environments.":[54],"We":[55],"present":[56],"cuSOAP,":[57],"GPU-accelerated":[59],"generator":[60],"atom-wise":[62],"SOAP":[63,248],"vectors":[64],"their":[66,83],"analytic":[67],"derivatives.":[68],"Built":[69],"on":[70,213,224],"PyTorch,":[71],"cuSOAP":[72],"evaluates":[73],"closed-form":[74],"expressions":[75],"or":[76],"quadratures":[77],"for":[78,86,114,156,250],"projection":[80],"coefficients":[81],"Cartesian":[84],"gradients":[85],"Gaussian-type-orbital":[87],"polynomial":[89],"bases,":[91],"through":[92,217],"fused":[93],"CUDA":[94],"Triton":[96],"kernels":[97],"that":[98],"eliminate":[99],"multi-gigabyte":[101],"intermediates":[102],"naive":[105],"tensor":[106],"formulation.":[107],"package":[109],"drop-in":[112],"replacement":[113],"CPU-based":[116],"reference":[117],"DScribe,":[118,169],"reproducing":[119],"constructor":[121],"signature,":[122],"feature":[123],"ordering,":[124],"output":[126],"to":[127,162,173,200,203],"within":[128,254],"${\\sim}10^{-6}$,":[129],"accepts":[131],"structures":[132],"directly":[133],"as":[134,195,240],"Atomistic":[135],"Simulation":[136],"Environment":[137],"(ASE)":[138],"Atoms":[139],"objects.":[140],"On":[141],"an":[142],"NVIDIA":[143],"Grace--Blackwell":[144],"(GB200)":[145],"node,":[146,230],"single":[148],"Blackwell":[149],"GPU":[150,215],"generates":[151],"full":[153],"descriptor-plus-Jacobian":[154],"workload":[155],"1000-molecule":[158],"water":[159,206],"cluster":[160],"up":[161,202],"two":[163],"orders":[164],"magnitude":[166],"faster":[167],"than":[168],"i.e.,":[170],"from":[171],"$13\\times$":[172],"$227\\times$":[174],"across":[175],"entire":[177],"$(n_{\\max},":[178],"l_{\\max})$":[179],"hyperparameter":[180],"map,":[181],"with":[182,186,235],"speedup":[184],"growing":[185],"band":[189],"limit":[190],"$l_{\\max}$.":[191],"Descriptor-only":[192],"generation":[193],"scales":[194],"$t":[196],"\\propto":[197],"n^{1.34}$,":[198],"close":[199],"linear,":[201],"million-atom":[205],"cluster,":[207],"which":[208],"processed":[210],"19.5s":[212],"and,":[216],"shared-memory":[219],"multiprocessing":[220],"driver,":[221],"5.7s":[223],"four":[226],"GPUs":[227],"demonstrating":[231],"$\\sim$90%":[232],"parallel":[233],"efficiency":[234],"no":[236],"sign":[237],"saturation":[239],"devices":[241],"are":[242],"added.":[243],"These":[244],"results":[245],"bring":[246],"on-the-fly":[247],"evaluation":[249],"large-scale":[251],"simulation":[253],"reach.":[255]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-09-10T00:00:00"}