[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86097-en":3,"doc-seo-86097-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},86097,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",6,"Technology","AAFLOW+ Stateful Operator Abstraction with Zero-Copy Distributed KV Cache Orchestration","Multi-agent LLM systems combine retrieval, planning, and reasoning but repeatedly recompute shared context because agent communication remains text-centric and relies on expensive prefill. The AAFLOW+ approach extends agentic workflow operators to treat the KV cache as a first-class distributed systems object. AAFLOW+ constructs communication-aware graphs and provides operators for KV materialization, transfer, fork, composition, and eviction, enabling zero-copy, transfer-aware reuse of long context. Results show up to 50.2× TTFT reduction, up to 7.63× lower multi-agent compute cost at 16-agent scale, 1.72–6.10× KV memory reduction, and over 7.74× throughput gain, supported by an analytical cost model grounded in hardware microbenchmarks.","[AAFLOW+] Stateful Operator Abstraction with Zero-Copy Distributed KV Cache Orchestration for Multi-Agent Workflows  \nArup Kumar Sarker Alexander James Halpern Mills Staylor  \nUniversity of Virginia, Biocomplexity Institute and Initiative Charlottesville, VA, USA [djy8hg@virginia.edu](djy8hg@virginia.edu)[ ](djy8hg@virginia.edu)[halperna22@gmail.com](halperna22@gmail.com)[ ](halperna22@gmail.com)[qad5gv@virginia.edu](qad5gv@virginia.edu)  \nGregor von Laszewski Geoffrey Fox Yue Cheng  \nBiocomplexity Institute and Initiative University of Virginia Charlottesville, VA, USA [laszewski@gmail.com](laszewski@gmail.com)[ ](laszewski@gmail.com)[vxj6mb@virginia.edu](vxj6mb@virginia.edu)[ ](vxj6mb@virginia.edu)[mrz7dp@virginia.edu](mrz7dp@virginia.edu)  \nAymen Alsaadi ShantenuJha  \nRutgers University Princeton Plasma Physics Laboratory Princeton, NJ, USA [aymen.alsaadi@rutgers.edu](aymen.alsaadi@rutgers.edu)[ ](aymen.alsaadi@rutgers.edu)[shantenu.jha@rutgers.edu](shantenu.jha@rutgers.edu)  \narXiv :2607 . 10987v1 [ cs .DC] 13 Jul 2026  \nABSTRACT  \nMulti-agent LLM systems increasingly integrate retrieval, planning, and reasoning, but remain fundamentally text-centric, requiring agents to repeatedly recompute shared context through expensive prefill. Although single-request inference is known tobe accelerated by KV-cache management, it is usually restricted to local serving scopes. We introduce AAFLOW+, a stateful extension of agentic workflow operators that makes KV cache a first-class distributed systems object. AAFLOW+ builds processes into communication-aware graphs that concurrently optimize data, prompts, and reusable model state. It also provides operators for KV materialization, transfer, fork, composition, and eviction. Its runtime enables zero-copy, transfer-aware execution, allowing agents to reuse long context without recomputation. AAFLOW+ reduces TTFT by up to 50.2×, achieves up to 7.63× reduced multi-agent compute cost at 16-agent scale, reduces KV memory by 1.72–6.10×, and increases throughput by more than 7.74×, based on an analytical cost model parameterized by empirical hardware microbenchmarks. The results demonstrate that KV transmission outperforms recomputation on networks with moderate to high bandwidth, making sure KV-state sharing greatly increases efficiency in multi-agent LLM systems by replacing text passing.  \nPVLDB Reference Format:  \nArup Kumar Sarker, Alexander James Halpern, Mills Staylor, Gregor von Laszewski, Geoffrey Fox, Yue Cheng, Aymen Alsaadi, and ShantenuJha.[AAFLOW+] Stateful Operator Abstraction with Zero-Copy Distributed KV Cache Orchestration for Multi-Agent Workflows. PVLDB, 14(1):  \nXXX-XXX, 2020 .  \ndoi:XX.XX/XXX.XX  \nPVLDB Artifact Availability:  \nThe source code, data, and/or other artifacts have been made available in AAFLOW/stateful_agentic_algebra directory at [https://github.com/aru](https://github.com/aru)pcsedu/AAFLOW.  \nThis work is licensed under the Creative Commons BY-NC-ND 4.0 International License. Visit [https://creativecommons.org/licenses/by-nc-nd/4.0/ to view a copy of](https://creativecommons.org/licenses/by-nc-nd/4.0/ to view a copy of)[ ](https://creativecommons.org/licenses/by-nc-nd/4.0/ to view a copy of)[this license. For any use beyond those covered by this license](this license. For any use beyond those covered by this license), [obtain permission by](obtain permission by)[emailing info@vldb.org. Copyright](emailing info@vldb.org. Copyright) is held by the owner/author(s). Publication rights licensed to the VLDB Endowment.  \nProceedings of the VLDB Endowment, Vol. 14, No. 1 ISSN 2150-8097 .  \ndoi:XX.XX/XXX.XX  \n1 INTRODUCTION  \nLarge language models are increasingly deployed as agentic systems that interleave retrieval, reasoning, tool invocation, and memory across multiple stages. Frameworks such as ReAct, Reflexion, AutoGen, and DSPy expand this design space, while RAG improves factual grounding by conditioning generation on external data [16, 20, 38, 41, 42] . How","cbCaigugfaWyzb67","https://ap.wps.com/l/cbCaigugfaWyzb67","pdf",902119,1,21,"English","en",105,"# Abstract\n# Introduction","[{\"question\":\"What problem does AAFLOW+ address in multi-agent LLM workflows?\",\"answer\":\"It addresses the inefficiency caused by text-centric agent communication, where shared context must be recomputed via costly prefill despite already being available.\"},{\"question\":\"How does AAFLOW+ improve KV-cache reuse across agents?\",\"answer\":\"It treats the KV cache as a distributed, first-class object and uses communication-aware graphs and operators for KV materialization, transfer, fork, composition, and eviction to enable zero-copy, transfer-aware reuse.\"},{\"question\":\"What performance benefits does the document report for AAFLOW+?\",\"answer\":\"It reports up to 50.2× TTFT reduction, up to 7.63× reduced multi-agent compute cost at 16-agent scale, 1.72–6.10× KV memory reduction, and more than 7.74× throughput improvement.\"}]",1784208490,53,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"aaflow-stateful-operator-abstraction-with-zero-copy-distributed-kv-cache-orchestration","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/aaflow-stateful-operator-abstraction-with-zero-copy-distributed-kv-cache-orchestration/86097/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does AAFLOW+ address in multi-agent LLM workflows?","Question",{"text":75,"@type":76},"It addresses the inefficiency caused by text-centric agent communication, where shared context must be recomputed via costly prefill despite already being available.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does AAFLOW+ improve KV-cache reuse across agents?",{"text":80,"@type":76},"It treats the KV cache as a distributed, first-class object and uses communication-aware graphs and operators for KV materialization, transfer, fork, composition, and eviction to enable zero-copy, transfer-aware reuse.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance benefits does the document report for AAFLOW+?",{"text":84,"@type":76},"It reports up to 50.2× TTFT reduction, up to 7.63× reduced multi-agent compute cost at 16-agent scale, 1.72–6.10× KV memory reduction, and more than 7.74× throughput improvement.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,113,118,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":114,"doc_module":4,"doc_module_name":45,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":45,"category_name":120,"show_sort_weight":121,"slug":122},8,"Research & Report",30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]