[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117532-en":3,"doc-seo-117532-105":30,"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":4,"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":27,"seo_description":14,"update_tm":28,"read_time":29},117532,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Benchmarking Machine Learning Models at the Edge - M.Sc. Thesis","Edge devices with limited resources have become a major research focus for running machine learning tasks closer to sensors and IoT systems. This M.Sc. thesis evaluates multiple ML models on a range of low- and high-end edge devices to determine suitability for deployment. The benchmarking pipeline is split into initial and execution phases, with rigorous timing and resource measurements to expose bottlenecks and quantify latency and speed. Warmup effects are highlighted, and the workflow is automated using Ansible to enable reproducible device benchmarking, with performance findings, limitations, and future research directions summarized.","Md Masud Haider  \nBENCHMARKING MACHINE LEARNING MODELS AT THE EDGE  \nM.Sc. Thesis  \nFaculty of Information Technology and Communication Sciences Supervisor: Kari Systä  \nExaminer: Jyke Savia November 2024  \nABSTRACT  \nAuthor: Md Haider  \nM.Sc. Thesis Tampere University  \nM.Sc. in Computing Sciences-Software, Web & Cloud November 2024  \nRecently the use of resource-constrained edge devices to perform machine learning tasks has gained substantial research attention. The need to perform such tasks arises from the vast amounts of sensors, and IoT and smart devices that generate large swaths of data at an enormous rate. To process such data in a real-time manner, cloud computing has proven to be suboptimal.  \nAs an alternative , edge devices paired with ML models can execute demanding tasks in diverse environments. This study explored a range of machine learning models coupled with several low and high-end edge devices to test their suitability for deployment in an edge environment. Unlike previous works, this experiment takes into account the initial heavy lifting that machine learning models and devices go through before performing the actual task of classification or detection. To systematically benchmark and gather results, this study divides the ML pipeline into initial and execution phases while thoroughly measuring each phase’s time and resource usage. This type of rigorous performance metric analysis enables decision-making about which machine learning model and devices have better suitability for latency and speed. Moreover, to add a device in the suit of existing ones and replicate the benchmark workflow, this study has automated the benchmark with the Ansible automation tool.  \nThe thesis research presents its findings from different devices. The division of phases allows for a clear understanding of the bottlenecks experienced by the devices and their capabilities. Moreover, an underlying crucial detail such as the warmup was revealed by scrutinizing the performances of the devices from the very beginning.  \nThe thesis concludes with performance highlights and limitations of the benchmark , identifying possible future research directions that can be developed further.  \nKeywords: Benchmark, Edge , ML Runtimes, GPU Acceleration, ML Performance , IoT, Machine Learning  \nThe originality of this thesis has been checked using the Turnitin Originality Check service.  \nUSE OF AI IN THESIS  \nI have utilised AI tools in my thesis:  \n☐ No  \n☒ Yes  \nThe AI tools utilised in my thesis and their purposes are described below:  \nNames and versions of AI tools:  \nChatGPT Free version: 4o  \nGrammarly Free version 1.2.115.1530  \nPurpose of using AI tools: [Provide a detailed explanation of the purpose and application of AI tools during your thesis process]  \nThe purpose was to check sentences for grammar and if they were correct and meaningful inan academic setting. Also, getting ideas about what kind of literature to cover around a certain topic. Grammarly claims its product uses AI to detect gaps in formulating coherent sentence writing. Grammarly was used only for grammar and phrase correctness checking , and its suggestions were taken where mistakes seemed to produce incorrect sentences. Summarizing a specific topic: For example, there are lot of information about runtimes used and even whole books exist about their functionalities and techniques used. So, a summary of a certain technology or tool was derived from an AI tool but not directly used.  \nSections where AI tools were used:  \nSection 1.1 Characteristics of ML Models Used  \nSection 1.2 and its subsections in a minimal way.  \nI acknowledge that I am fully responsible for the entire content of my thesis, including the parts generated by AI, and accept accountability for any violations of ethical standards in publications.  \nPREFACE  \nThis thesis titled ‘Benchmarking Machine Learning Models at the Edge’, contributes to the field of machine learning and edge computing as part of","cbCais2GCxt6d9AV","https://ap.wps.com/l/cbCais2GCxt6d9AV","pdf",2405977,1,81,"English","en",105,"# Introduction\n# Literature Review\n## Benchmarking in the Context of Machine Learning\n## Characteristics of ML Models Used\n## Fundamentals of Cloud and Edge\n### Cloud Computing\n### Edge Computing\n### Fog Computing\n### Edge AI\n### Cloud-Edge-Fog Collaboration\n# Methodology and Benchmarking Pipeline\n## Initial and Execution Phases\n## Performance Metrics and Measurements\n# Experiment Setup and Devices\n# Results and Discussion\n# Conclusion and Future Work","[{\"question\":\"What is the main research goal of the thesis?\",\"answer\":\"To benchmark different machine learning models running on various edge devices and determine which combinations are better suited for latency and speed under edge deployment constraints.\"},{\"question\":\"How does the thesis design its benchmarking methodology?\",\"answer\":\"It divides the ML pipeline into initial and execution phases and measures time and resource usage for each phase to isolate bottlenecks.\"},{\"question\":\"Why is warmup emphasized in the performance evaluation?\",\"answer\":\"Warmup is identified by scrutinizing device performance from the very beginning, revealing a crucial factor that affects observed results.\"}]","Benchmarking Machine Learning Models at the Edge - M.Sc. Thesis | PDF",1785676732,204,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"benchmarking-machine-learning-models-at-the-edge-msc-thesis","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/benchmarking-machine-learning-models-at-the-edge-msc-thesis/117532/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main research goal of the thesis?","Question",{"text":75,"@type":76},"To benchmark different machine learning models running on various edge devices and determine which combinations are better suited for latency and speed under edge deployment constraints.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the thesis design its benchmarking methodology?",{"text":80,"@type":76},"It divides the ML pipeline into initial and execution phases and measures time and resource usage for each phase to isolate bottlenecks.",{"name":82,"@type":73,"acceptedAnswer":83},"Why is warmup emphasized in the performance evaluation?",{"text":84,"@type":76},"Warmup is identified by scrutinizing device performance from the very beginning, revealing a crucial factor that affects observed results.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]