Filters








3,703 Hits in 5.6 sec

SAMURAI: A Streaming Multi-tenant Context-Management Architecture for Intelligent and Scalable Internet of Things Applications

Davy Preuveneers, Yolande Berbers
2014 2014 International Conference on Intelligent Environments  
We present SAMURAI, a multi-tenant streaming context architecture that integrates and exposes well-known components for complex event processing, machine learning, knowledge representation, NoSQL persistence  ...  In the Internet of Things, heterogeneous and distributed streams of sensor events is a driver for contextaware behavior in intelligent environments.  ...  A limitation of our current experiments was that the load generation for simulation (external) and the multi-tenant distributed deployment of SAMURAI (internal) were all linked to the same local network  ... 
doi:10.1109/ie.2014.43 dblp:conf/intenv/PreuveneersB14 fatcat:w6d5623fcvgbdomurg5af5szge

No DNN Left Behind: Improving Inference in the Cloud with Multi-Tenancy [article]

Amit Samanta and Suhas Shrinivasan and Antoine Kaufmann and Jonathan Mace
2019 arXiv   pre-print
With the rise of machine learning, inference on deep neural networks (DNNs) has become a core building block on the critical path for many cloud applications.  ...  We argue that DNN inference is an ideal candidate for a multi-tenant system because of its narrow and well-defined interface and predictable resource requirements.  ...  Motivation Deep neural networks (DNNs) excel at a wide range of machine learning tasks including computer vision, natural language processing, speech detection, and more.  ... 
arXiv:1901.06887v2 fatcat:k6bgjy7m3vcvrog5skrh55bzza

Delivering a machine learning course on HPC resources

Federica Legger, Stefano Bagnasco, Stefano Lusso, Gabriele Gaetano Fronze', Sara Vallero
2019 Zenodo  
Machine learning algorithms typically require large sets of data to train the models and extensive usage of computing resources both for training and inference.  ...  GPUs) resources distributed over different hosts across a network.  ...  "Searching for Exotic Particles in High-energy Physics with Deep Learning."  ... 
doi:10.5281/zenodo.3599628 fatcat:nlhsstfzzrhx7dpjruf5li7tkq

SESAME: Software defined Enclaves to Secure Inference Accelerators with Multi-tenant Execution [article]

Sarbartha Banerjee, Prakash Ramrakhyani, Shijia Wei, Mohit Tiwari
2020 arXiv   pre-print
learning models.  ...  specific defenses ranging from 3.96% to 34.87% (across confidential inputs and models and single vs. multi-tenant systems).  ...  Specifically, we introduce Sesame, a software-defined enclave framework for multi-tenant machine learning inference accelerators that are tightly coupled to a CPU (e.g., Arm Ethos-N NPUs [5, 17] ).  ... 
arXiv:2007.06751v2 fatcat:e2nq6hsp4rbi5p4mn3v27lecfu

AI on the Edge: Rethinking AI-based IoT Applications Using Specialized Edge Architectures [article]

Qianlin Liang, Prashant Shenoy, David Irwin
2020 arXiv   pre-print
We find that edge accelerators can support varying degrees of concurrency for multi-tenant inference applications, but lack isolation mechanisms necessary for edge cloud multi-tenant hosting.  ...  The attractiveness of edge computing has been further enhanced due to the recent availability of special-purpose hardware to accelerate specific compute tasks, such as deep learning inference, on edge  ...  multi-tenant hosting.  ... 
arXiv:2003.12488v1 fatcat:rice6s77jjevlk3ir4em6doc2e

Deep-Dup: An Adversarial Weight Duplication Attack Framework to Crush Deep Neural Network in Multi-Tenant FPGA [article]

Adnan Siraj Rakin, Yukui Luo, Xiaolin Xu, Deliang Fan
2021 arXiv   pre-print
Such a multi-tenant FPGA setup for DNN acceleration potentially exposes DNN interference tasks under severe threat from malicious users.  ...  The wide deployment of Deep Neural Networks (DNN) in high-performance cloud computing platforms brought to light multi-tenant cloud field-programmable gate arrays (FPGA) as a popular choice of accelerator  ...  With Machine Learning as a service (MLaaS) [40, 41] becoming popular, public lease FPGAs also become an emerging platform for acceleration purposes.  ... 
arXiv:2011.03006v2 fatcat:rn3n2fian5fllkdv2k2y75c7s4

SoK: On the Security Challenges and Risks of Multi-Tenant FPGAs in the Cloud [article]

Shaza Zeitouni, Ghada Dessouky, Ahmad-Reza Sadeghi
2020 arXiv   pre-print
In their continuous growth and penetration into new markets, Field Programmable Gate Arrays (FPGAs) have recently made their way into hardware acceleration of machine learning among other specialized compute-intensive  ...  In this paper, we survey industrial and academic deployment models of multi-tenant FPGAs in the cloud computing settings, and highlight their different adversary models and security guarantees, while shedding  ...  The AaaS model provides tenants with FPGA-accelerated services, where the FPGA is deployed to accelerate a specific pre-defined functionality, such as a machine learning computation.  ... 
arXiv:2009.13914v2 fatcat:mbdpjfuoljderjhopoppxkxkoe

A Unified FPGA Virtualization Framework for General-Purpose Deep Neural Networks in the Cloud

Shulin Zeng, Guohao Dai, Hanbo Sun, Jun Liu, Shiyao Li, Guangjun Ge, Kai Zhong, Kaiyuan Guo, Yu Wang, Huazhong Yang
2022 ACM Transactions on Reconfigurable Technology and Systems  
Aiming to solve these problems, we propose a unified virtualization framework for general-purpose deep neural networks in the cloud, enabling multi-tenant sharing for both the Convolution Neural Network  ...  On the other hand, current cloud-based DNN accelerators have excessive compilation overhead, especially when scaling out to multi-FPGA systems for multi-tenant sharing, leading to unacceptable compilation  ...  In this article, we enable the virtualization on a single FPGA of the node level, where a single-node multi-tenant DNN accelerator based on FPGA is also enabled for deep learning inference applications  ... 
doi:10.1145/3480170 fatcat:kgrhiohisvcdxm635l3wykgdri

FPGA-Accelerated Analytics: From Single Nodes to Clusters

Zsolt István, Kaan Kara, David Sidler
2020 Foundations and Trends in Databases  
, Machine Learning and Distributed Joins.  ...  As a prominent example, the Microsoft Catapult project uses FPGAs to create programmable network-interface cards (NICs) to offload tenant network functions from the CPU.  ... 
doi:10.1561/1900000072 fatcat:ghwjkt7yh5c2vpvmilxbh2a534

The Coming Age of Pervasive Data Processing

Jan S. Rellermeyer, Sobhan Omranian Khorasani, Dan Graur, Apourva Parthasarathy
2019 2019 18th International Symposium on Parallel and Distributed Computing (ISPDC)  
In machine learning, the use of GPUs as accelerators is by now best practice.  ...  Spark-GPU [61] accelerates Spark workloads through GPUs and reports a speedup of 16.13x for machine learning workloads and 4.83x for SQL queries.  ... 
doi:10.1109/ispdc.2019.00011 dblp:conf/ispdc/RellermeyerKGP19 fatcat:kczllhax3rgkbeuhmblp6wsowq

Large-Scale Intelligent Microservices [article]

Mark Hamilton, Nick Gonsalves, Christina Lee, Anand Raman, Brendan Walsh, Siddhartha Prasad, Dalitso Banda, Lucy Zhang, Mei Gao, Lei Zhang, William T. Freeman
2021 arXiv   pre-print
Deploying Machine Learning (ML) algorithms within databases is a challenge due to the varied computational footprints of modern ML algorithms and the myriad of database technologies each with its own restrictive  ...  To eliminate the majority of overhead from network communication, we also introduce a low-latency containerized version of our architecture.  ...  Furthermore we thank the CSAIL Alliances Systems that Learn program for helping to fund this work.  ... 
arXiv:2009.08044v3 fatcat:umyfqotpajgb5fdty6pscs5vhy

LUMEN: A global fault management framework for network virtualization environments

Sihem Cherrared, Sofiane Imadali, Eric Fabre, Gregor Goessler
2018 2018 21st Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN)  
However, it introduces new fault management challenges including dynamic topology, multi-tenant fault isolation and data consistency and ambiguity; that we propose to define in this paper.  ...  LUMEN includes the canonical steps of the fault management process and proposes a monitoring solution for all types of Network virtualization Environments.  ...  Acknowledgment Authors would like to thank their colleague, Ayoub Bousselmi, for his early review of the paper.  ... 
doi:10.1109/icin.2018.8401622 dblp:conf/icin/CherraredIFG18 fatcat:mzuk2f6oibbyfktoauq4zv54ca

Analysis of Large-Scale Multi-Tenant GPU Clusters for DNN Training Workloads [article]

Myeongjae Jeon, Shivaram Venkataraman, Amar Phanishayee, Junjie Qian, Wencong Xiao, Fan Yang
2019 arXiv   pre-print
With widespread advances in machine learning, a number of large enterprises are beginning to incorporate machine learning models across a number of products.  ...  These models are typically trained on shared, multi-tenant GPU clusters.  ...  Acknowledgments We thank our shepherd, David Nellans, and the anonymous reviewers for their valuable comments and suggestions.  ... 
arXiv:1901.05758v2 fatcat:fp3vmcibtngwxejceaztiwrrem

Lazy Ctrl: Scalable Network Control for Cloud Data Centers

Lin Wang, Kai Zheng, Baohua Yang, Yi Sun, Yue Zhang, Steve Uhlig
2015 2015 IEEE 35th International Conference on Distributed Computing Systems  
To address this, we present LazyCtrl, a novel hybrid control plane design for data center networks where network control is carried out by distributed control mechanisms inside independent groups of switches  ...  The advent of software defined networking enables flexible, reliable and feature-rich control planes for data center networks.  ...  Relatively Stable Tenant Size For multi-tenant cloud data centers, we observe that the number of virtual machines for a single tenant is changing slightly, while the number of tenant users, as well as  ... 
doi:10.1109/icdcs.2015.110 dblp:conf/icdcs/WangZYSZU15 fatcat:2xsizicg35cdpeipmg6ao2ift4

LazyCtrl: Scalable Network Control for Cloud Data Centers [article]

Kai Zheng, Lin Wang, Baohua Yang, Yi Sun, Yue Zhang, Steve Uhlig
2015 arXiv   pre-print
To address this, we present LazyCtrl, a novel hybrid control plane design for data center networks where network control is carried out by distributed control mechanisms inside independent groups of switches  ...  The advent of software defined networking enables flexible, reliable and feature-rich control planes for data center networks.  ...  Relatively Stable Tenant Size For multi-tenant cloud data centers, we observe that the number of virtual machines for a single tenant is changing slightly, while the number of tenant users, as well as  ... 
arXiv:1504.02609v1 fatcat:wzzd5i6bsnaldlmug5jg3rkptu
« Previous Showing results 1 — 15 out of 3,703 results