Filters








70,984 Hits in 4.2 sec

A Power Benchmarking Framework for Network Devices [chapter]

Priya Mahadevan, Puneet Sharma, Sujata Banerjee, Parthasarathy Ranganathan
2009 Lecture Notes in Computer Science  
Researchers have proposed several strategies for energy management of networking devices.  ...  We build and describe a benchmarking suite that will allow users to measure and compare the power consumed for a large set of common configurations at any switch or router of their choice.  ...  Benchmarking Framework We begin by describing factors that are commonly used and most likely to affect a switch or router's power consumption.  ... 
doi:10.1007/978-3-642-01399-7_62 fatcat:unri2wlaargsvftua4ptv5hyqy

Performance of deep neural networks on low-power IoT devices

Christos Profentzas, Magnus Almgren, Olaf Landsiedel
2021 Proceedings of the Workshop on Benchmarking Cyber-Physical Systems and Internet of Things  
However, running deep neural networks on low-power IoT devices is challenging due to their resource-constraints in memory, compute power, and energy.  ...  Our benchmark reveals significant differences and trade-offs for each framework and its toolchain: (1) We find that uTensor is the most straightforward framework to use, followed by TF-Micro, and then  ...  We present a benchmark to evaluate three representative frameworks for DNNs inference on low-power IoT devices.  ... 
doi:10.1145/3458473.3458823 fatcat:yacip4x67ffwxg2lywnj5iq2qi

Benchmarking Contemporary Deep Learning Hardware and Frameworks:A Survey of Qualitative Metrics [article]

Wei Dai, Daniel Berleant
2019 arXiv   pre-print
This paper surveys benchmarking principles, machine learning devices including GPUs, FPGAs, and ASICs, and deep learning software frameworks.  ...  Because MLPerf is a benchmark organization working with industry and academia, and offering deep learning benchmarks that evaluate training and inference on deep learning hardware devices, the survey also  ...  In this paper we introduce 11 qualitative benchmarking metrics for hardware devices and six metrics for software frameworks in deep learning, respectively.  ... 
arXiv:1907.03626v4 fatcat:cabjqm655bcw5euk2hessp4axy

Benchmarking Contemporary Deep Learning Hardware and Frameworks: A Survey of Qualitative Metrics

Wei Dai, Daniel Berleant
2019 2019 IEEE First International Conference on Cognitive Machine Intelligence (CogMI)  
This paper surveys benchmarking principles, machine learning devices including GPUs, FPGAs, and ASICs, and deep learning software frameworks.  ...  Because MLPerf is a benchmark organization working with industry and academia, and offering deep learning benchmarks that evaluate training and inference on deep learning hardware devices, the survey also  ...  In this paper we introduce 11 qualitative benchmarking metrics for hardware devices and six metrics for software frameworks in deep learning, respectively.  ... 
doi:10.1109/cogmi48466.2019.00029 dblp:conf/cogmi/DaiB19 fatcat:3tziv3stuvddnetxc6ha66bcgq

TinyML Platforms Benchmarking [article]

Anas Osman, Usman Abid, Luca Gemma, Matteo Perotto, Davide Brunelli
2021 arXiv   pre-print
TinyML provides a unique solution by aggregating and analyzing data at the edge on low-power embedded devices.  ...  Recent advances in state-of-the-art ultra-low power embedded devices for machine learning (ML) have permitted a new class of products whose key features enable ML capabilities on microcontrollers with  ...  This section outlines the benchmarking setting for our two use cases. Each benchmark targets a specific use case with a different dataset, modelled on two separate targets.  ... 
arXiv:2112.01319v1 fatcat:lhaxtpg5g5c3vmzdfxoh5robqu

Computation Offloading from Mobile Devices

Arani Bhattcharya, Pradipta De
2016 Proceedings of the Third International Workshop on Adaptive Resource Management and Scheduling for Cloud Computing - ARMS-CC'16  
In this paper, we determine the devices whose processors have sufficient power to act as servers for computation offloading.  ...  However, offloading to cloud data centers has a high network latency.  ...  The method Latency Processor Power Edge Device 10 ms 2 times Cloud Server 50 ms 10 times (a) Parameters used for the example. Latency refers to the latency of server from mobile device.  ... 
doi:10.1145/2962564.2962569 dblp:conf/podc/BhattacharyaD16 fatcat:ih53byryx5b43ilkppiysryxu4

MAMoC: Multisite Adaptive Offloading Framework for Mobile Cloud Applications [article]

Dawand Sulaiman, Adam Barker
2017 arXiv   pre-print
This paper presents MAMoC, a framework which brings together a diverse range of infrastructure types including mobile devices, cloudlets, and remote cloud resources under one unified API.  ...  The results show that offloading computation using our framework can reduce the overall task completion time for both single-site and multi-site offloading scenarios.  ...  result of FFT B Average Benchmark workload result B mob Benchmark score of a mobile device B clet Benchmark score of a cloudlet Bc Benchmark score of a remote cloud BL mob Battery level of a  ... 
arXiv:1711.05518v1 fatcat:acur47gubjdwvfg4izggq3a2fq

WURBench: Toward Benchmarking Wake-up Radio-based Systems [article]

Rajeev Piyare, Amy L. Murphy
2018 arXiv   pre-print
This paper leads toward an evaluation framework, a benchmark, to enable accurate and repeatable profiling of WUR-based systems, leading to more consistent and therefore comparable evaluations for current  ...  Standard methodologies for benchmarking are crucial for quantitatively evaluating the performance of this emerging technology, however, currently, no accepted standard for such quantitative measurement  ...  WUR hardware designers can also utilize this framework to benchmark devices against competitors. (iii) facilitate a repeatable test environment for WUR-based systems. II.  ... 
arXiv:1811.06890v1 fatcat:p65pjop7ivfebni6gqoe265tva

Process migration-based computational offloading framework for IoT-supported mobile edge/cloud computing

Abdullah Yousafzai, Ibrar Yaqoob, Muhammad Imran, Abdullah Gani, Rafidah Md Noor
2019 IEEE Internet of Things Journal  
In this paper, we analyze the effect of platform-dependent native applications on computational offloading in edge networks and propose a lightweight process migration-based computational offloading framework  ...  Hence, the proposed framework shows profound potential for resource-intensive IoT application processing in MEC.  ...  Similarly, a general framework for IoT fog-cloud applications, along with a delay-minimizing collaboration and offloading policy for fog-capable devices, was proposed in [18] .  ... 
doi:10.1109/jiot.2019.2943176 fatcat:7to2d5aqnndthjvx47bzluonoe

MAMoC: Multisite Adaptive Offloading Framework for Mobile Cloud Applications

Dawand Sulaiman, Adam Barker
2017 2017 IEEE International Conference on Cloud Computing Technology and Science (CloudCom)  
This paper presents MAMoC, a framework which brings together a diverse range of infrastructure types including mobile devices, cloudlets, and remote cloud resources under one unified API.  ...  The results show that offloading computation using our framework can reduce the overall task completion time for both single-site and multi-site offloading scenarios.  ...  result of FFT B Average Benchmark workload result B mob Benchmark score of a mobile device B clet Benchmark score of a cloudlet Bc Benchmark score of a remote cloud BL mob Battery level of a  ... 
doi:10.1109/cloudcom.2017.34 dblp:conf/cloudcom/SulaimanB17 fatcat:5cz6ubnunfho3mlpenysz3v2xm

Energy Audit: Monitoring power consumption in diverse network environments

Joseph Chabarek, Paul Barford
2013 2013 International Green Computing Conference Proceedings  
In this paper we present a tool suite called EnergyAudit that is designed for network operators to audit the power consumption in their infrastructure.  ...  The auditing tool can infer missing data and reports both the estimated power consumption for devices in the network and the fidelity of the measurements from which the device-to-power mapping was reported  ...  Power benchmarking frameworks for networking devices based on direct measurement are described in [23] , [26] .  ... 
doi:10.1109/igcc.2013.6604505 dblp:conf/green/ChabarekB13 fatcat:ezoutvbkxfgprniwwcdlk6ucwe

Tango: A Deep Neural Network Benchmark Suite for Various Accelerators [article]

Aajna Karki, Chethan Palangotu Keshava, Spoorthi Mysore Shivakumar, Joshua Skow, Goutam Madhukeshwar Hegde, Hyeran Jeon
2019 arXiv   pre-print
Though a few DNN benchmark suites have been recently released, most of them require to install proprietary DNN libraries or resource-intensive DNN frameworks, which are hard to run on resource-limited  ...  To provide more efficient computing platforms for DNN applications, it is essential to have evaluation environments that include assorted benchmark workloads.  ...  For the power measurement, we used GPUWattch [38] for detailed statistics and Wattsup power meter for device-level statistics.  ... 
arXiv:1901.04987v1 fatcat:5kcbyowrrngvrf2o3ks7rfklsm

Comparison and Benchmarking of AI Models and Frameworks on Mobile Devices [article]

Chunjie Luo, Xiwen He, Jianfeng Zhan, Lei Wang, Wanling Gao, Jiahui Dai
2020 arXiv   pre-print
Each network is implemented by three frameworks which are designed for mobile and embedded devices: Tensorflow Lite, Caffe2, Pytorch Mobile.  ...  Considering the model diversity and framework diversity, we propose a benchmark suite, AIoTBench, which focuses on the evaluation of the inference abilities of mobile and embedded devices.  ...  With diverse and representative models and frameworks, the mobile devices can get a comprehensive benchmarking and evaluation.  ... 
arXiv:2005.05085v1 fatcat:72p3whrkfvhvzpl6ih7pfc2bi4

A Data Stream Processing Optimisation Framework for Edge Computing Applications

Gayashan Amarasinghe, Marcos D. de Assuncao, Aaron Harwood, Shanika Karunasekera
2018 2018 IEEE 21st International Symposium on Real-Time Distributed Computing (ISORC)  
and network latency for a given data stream topology allocation and power consumption data.  ...  We use these micro-benchmarks to establish a baseline for our optimisation framework. As the second category we use a realistic edge-cloud application topology.  ... 
doi:10.1109/isorc.2018.00020 dblp:conf/isorc/AmarasingheAHK18 fatcat:5lejbbb3znaupnzxbmt6gsgbrq

TinyML for Ubiquitous Edge AI [article]

Stanislava Soro
2021 arXiv   pre-print
TinyML addresses the challenges in designing power-efficient, compact deep neural network models, supporting software framework, and embedded hardware that will enable a wide range of customized, ubiquitous  ...  devices operating at extremely low power range (mW range and below).  ...  Similar to MobileNet [1] , which became a baseline model for benchmarking different neural network models for mobile edge computing devices, a set of generally applicable TinyML-based learning models  ... 
arXiv:2102.01255v1 fatcat:if5ny6kcirdkhnj56mswfaptlm
« Previous Showing results 1 — 15 out of 70,984 results