[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123667-en":3,"doc-seo-123667-105":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},123667,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","Exploring the Impact of Serverless Computing on Peer To Peer Training - Machine Learning","The document addresses the growing need for computational power in big data and machine learning, where distributed training improves scalability and fault tolerance but also increases resource consumption, costs, and communication overhead as peers scale. It proposes a new architecture that integrates serverless computing with peer-to-peer networks, along with a method for efficient parallel gradient computation under resource constraints. Results show up to a 97.34% faster gradient computation time, while cost may rise to 5.4× compared with instance-based approaches. The pay-as-you-go model and dynamic resource allocation still make serverless promising for resource-limited ML workloads.","Exploring the Impact of Serverless Computing on Peer To Peer Training Machine Learning  \nAmine BARRAK∗ , Ranim TRABELSI∗ , Fehmi JAAFAR∗ , Fabio PETRILLO†  \n∗ Department of Computer Science and Mathematics, University of Quebec at Chicoutimi, UQAC, Saguenay, Canada  \nEmail: {mabarrak, ranim.trabelsi1, [fehmi.jaafar](fehmi.jaafar}@uqac.ca)[}](fehmi.jaafar}@uqac.ca)[@uqac.ca](fehmi.jaafar}@uqac.ca)[ ](fehmi.jaafar}@uqac.ca)†De´partement de ge´nie logiciel, ´Ecole de Technologie Sup e´rieure, ´ETS, Montreal, QC  \nEmail: [fabio.petrillo@etsmtl.ca](fabio.petrillo@etsmtl.ca)  \narXiv :2309 . 14139v1 [ cs .DC] 25 Sep 2023  \nAbstract—  \nThe increasing demand for computational power in big data and machine learning has driven the development of distributed training methodologies. Among these, peer-topeer (P2P) networks provide advantages such as enhanced scalability and fault tolerance. However, they also encounter challenges related to resource consumption, costs, and communication overhead as the number of participating peers grows. In this paper, we introduce a novel architecture that combines serverless computing with P2P networks for distributed training and present a method for efficient parallel gradient computation under resource constraints.  \nOur findings show a significant enhancement in gradient computation time, with up to a 97.34% improvement compared to conventional P2P distributed training methods. As for costs, our examination confirmed that the serverless architecture could incur higher expenses, reaching up to 5.4 times more than instance-based architectures. It is essential to consider that these higher costs are associated with marked improvements in computation time, particularly under resource-constrained scenarios.  \nDespite the cost-time trade-off, the serverless approach still holds promise due to its pay-as-you-go model. Utilizing dynamic resource allocation, it enables faster training times and optimized resource utilization, making it a promising candidate for a wide range of machine learning applications.  \nIndex Terms—Serverless, FaaS, Function as a Service, P2P, peer-to-peer architecture, Distributed Training, Machine Learning.  \nI. INTRODUCTION  \nThe exponential growth of data in the modern digital age [1] has transformed the landscape of artificial intelligence (AI) and machine learning (ML), propelling these fields into a new era of innovation and discovery. This vast deluge of data, has given rise to increasingly sophisticated and complex models that can extract valuable insights and make accurate predictions [2] . However, these sophisticated models pose a formidable challenge, due to the need for vast computational resources. This escalating demand for computational power has led to the emergence of distributed training [3] . By harnessing the combined power of multiple devices, the training methodology encompasses the division of the dataset among a cohort of  \nworkers, each training their local model replicas in parallel and iteratively. To ensure convergence, the workers periodically synchronize their updated local models [4] .  \nVarious topologies have been proposed in the literature [2] to facilitate distributed training, including parameter server [5]–[7] and peer-to-peer architectures [8]–[12] . In the parameter server architecture, the worker nodes perform computationson their respective data partitions and communicate with the parameter server to update the global model. In contrast, peerto-peer (P2P) architectures distribute the model parameters and computation across all nodes in the network, eliminating the need for a central coordinator [2] .  \nRegardless of the topology employed for distributed training, developers often struggle with managing resources and navigating the complexities of ML training. This can result in over-provisioning and diminished productivity, posing challenges for ML users striving to achieve optimal outcomes [13] .  \nTo address these challenges, building machine","cbCaipYnCD9nkzKJ","https://ap.wps.com/l/cbCaipYnCD9nkzKJ","pdf",774637,1,12,"English","en",105,"# Abstract\n# Index Terms\n# Introduction","[{\"question\":\"What problem does the document address in distributed machine learning?\",\"answer\":\"It addresses the escalating computational demand that drives distributed training, along with the growing challenges of resource consumption, costs, and communication overhead as the number of peers increases.\"},{\"question\":\"What approach is proposed for serverless-enabled peer-to-peer distributed training?\",\"answer\":\"It introduces an architecture combining serverless computing with peer-to-peer networks and a method for efficient parallel gradient computation while operating under resource constraints.\"},{\"question\":\"What are the reported performance and cost outcomes?\",\"answer\":\"The method achieves up to a 97.34% improvement in gradient computation time versus conventional P2P distributed training, but serverless can cost up to 5.4 times more than instance-based architectures.\"}]","Exploring the Impact of Serverless Computing on Peer To Peer Training - Machine Learning | PDF",1785817914,30,{"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},"exploring-the-impact-of-serverless-computing-on-peer-to-peer-training-machine-learning","",{"@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/exploring-the-impact-of-serverless-computing-on-peer-to-peer-training-machine-learning/123667/",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","2026-08-04",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 problem does the document address in distributed machine learning?","Question",{"text":75,"@type":76},"It addresses the escalating computational demand that drives distributed training, along with the growing challenges of resource consumption, costs, and communication overhead as the number of peers increases.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What approach is proposed for serverless-enabled peer-to-peer distributed training?",{"text":80,"@type":76},"It introduces an architecture combining serverless computing with peer-to-peer networks and a method for efficient parallel gradient computation while operating under resource constraints.",{"name":82,"@type":73,"acceptedAnswer":83},"What are the reported performance and cost outcomes?",{"text":84,"@type":76},"The method achieves up to a 97.34% improvement in gradient computation time versus conventional P2P distributed training, but serverless can cost up to 5.4 times more than instance-based architectures.","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},[93,97,101,105,110,115,120,122,127,130,134],{"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":29,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]