Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Filters








408 Hits in 4.5 sec

An experimental evaluation of datacenter workloads on low-power embedded micro servers

Yiran Zhao, Shen Li, Shaohan Hu, Hongwei Wang, Shuochao Yao, Huajie Shao, Tarek Abdelzaher
2016 Proceedings of the VLDB Endowment  
This paper presents a comprehensive evaluation of an ultralow power cluster, built upon the Intel Edison based micro servers.  ...  Although sensor-class micro servers have much less capacity, they are touted for minimal power consumption (< 1 Watt), which opens new possibilities of achieving higher energy efficiency in datacenter  ...  More recently, the idea of building datacenters based on low-power embedded systems has become popular.  ... 
doi:10.14778/2947618.2947625 fatcat:fzqtqu55erafpjyvlmcc67u4fm

An energy case for hybrid datacenters

Byung-Gon Chun, Gianluca Iannaccone, Giuseppe Iannaccone, Randy Katz, Gunho Lee, Luca Niccolini
2010 ACM SIGOPS Operating Systems Review  
To address this challenge, we explore the potential of hybrid datacenter designs that mix low power platforms with high performance ones.  ...  We evaluate the feasibility of our approach through experiments and then discuss the design challenges and options of hybrid datacenters.  ...  Lim et. al [18] evaluated an alternative server architecture design built using embedded components and showed improvement in performance per dollar.  ... 
doi:10.1145/1740390.1740408 fatcat:7zfwk6jf3jgubjyhcl2bfij3zy

BAAT: Towards Dynamically Managing Battery Aging in Green Datacenters

Longjun Liu, Chao Li, Hongbin Sun, Yang Hu, Juncheng Gu, Tao Li
2015 2015 45th Annual IEEE/IFIP International Conference on Dependable Systems and Networks  
To fill this critical void, we thoroughly investigate battery aging on a heavily instrumented prototype over an observation period of six months.  ...  Energy storage devices (batteries) have shown great promise in eliminating supply/demand power mismatch and reducing energy/power cost in green datacenters.  ...  Battery management in micro-systems: As the primary power source in micro-systems, battery usage and lifetime management have been studied in light of mobile/embedded environment [45] [46] [47] [48]  ... 
doi:10.1109/dsn.2015.43 dblp:conf/dsn/LiuLSHGL15 fatcat:sfgjapjlbfagxnjdyslgo3aowq

A performance study of big data on small nodes

Dumitrel Loghin, Bogdan Marius Tudor, Hao Zhang, Beng Chin Ooi, Yong Meng Teo
2015 Proceedings of the VLDB Endowment  
Traditionally, datacenters employ x86-64 (big) server nodes with power usage of tens to hundreds of Watts.  ...  We evaluate execution time, energy usage and total cost of running the workloads on selfhosted ARM and Xeon nodes.  ...  This work shows that highend notebooks with Intel Core processors are 300% and 80% more energy-efficient than low-power server systems and low-power embedded systems, respectively.  ... 
doi:10.14778/2752939.2752945 fatcat:utk3f3g4ubhcpdh46fk6nwtxty

Evaluating the Combined Impact of Node Architecture and Cloud Workload Characteristics on Network Traffic and Performance/Cost

Diman Zad Tootaghaj, Farshid Farhat, Mohammad Arjomand, Paolo Faraboschi, Mahmut Taylan Kandemir, Anand Sivasubramaniam, Chita R. Das
2015 2015 IEEE International Symposium on Workload Characterization  
Motivated by this observation, this paper performs a thorough characterization of twelve cloud workloads using a full-system datacenter simulation infrastructure.  ...  The combined impact of node architecture and workload characteristics on off-chip network traffic with performance/cost analysis has not been investigated before in the context of emerging cloud applications  ...  Based on our experimental results, one can reuse the results to analyze similar studied workloads.  ... 
doi:10.1109/iiswc.2015.31 dblp:conf/iiswc/TootaghajFAFKSD15 fatcat:s3dgaxjr6vdsfm5fld4zzjmlg4

SWEEP

Kristof Du Bois, Tim Schaeps, Stijn Polfliet, Frederick Ryckbosch, Lieven Eeckhout
2011 Proceedings of the 6th International Conference on High Performance and Embedded Architectures and Compilers - HiPEAC '11  
The focus on energy efficiency has led to a number of power benchmarking methods recently.  ...  We employ SWEEP to generate a wide range of synthetic workloads while varying the instruction mix, ILP, memory access patterns, and I/O-intensiveness; and we use SWEEP to evaluate the energy efficiency  ...  INTRODUCTION Energy efficiency has emerged as a primary design concern across the entire compute range, from low-end embedded systems to highend servers and datacenters.  ... 
doi:10.1145/1944862.1944886 dblp:conf/hipeac/BoisSPRE11 fatcat:kies6m7pqvdnhfkgtox7ivkjaq

GreenHadoop

Íñigo Goiri, Kien Le, Thu D. Nguyen, Jordi Guitart, Jordi Torres, Ricardo Bianchini
2012 Proceedings of the 7th ACM european conference on Computer Systems - EuroSys '12  
Interest has been growing in powering datacenters (at least partially) with renewable or "green" sources of energy, such as solar or wind.  ...  We propose GreenHadoop, a MapReduce framework for a datacenter powered by a photovoltaic solar array and the electrical grid (as a backup).  ...  We are also grateful to our sponsors: Spain's Ministry of Science and Technology and the European Union under contract TIN2007-60625, COST Action IC0804 and grant AP2008-0264, the Generalitat de Catalunya  ... 
doi:10.1145/2168836.2168843 dblp:conf/eurosys/GoiriLNGTB12 fatcat:crxyh7hes5gylbopvf4fet3uhe

Thermal time shifting

Matt Skach, Manish Arora, Chang-Hong Hsu, Qi Li, Dean Tullsen, Lingjia Tang, Jason Mars
2015 SIGARCH Computer Architecture News  
We present and validate a methodology to study the impact of PCM on a datacenter, and evaluate two important opportunities for cost savings.  ...  However, this growth comes at the cost of an increasing thermal load that must be removed to prevent overheating and server failure.  ...  AMD, the AMD Arrow logo, and combinations thereof are trademarks of Advanced Micro Devices, Inc.  ... 
doi:10.1145/2872887.2749474 fatcat:qi73vgmjnzcw7djjtscylmtspy

Thermal time shifting

Matt Skach, Manish Arora, Chang-Hong Hsu, Qi Li, Dean Tullsen, Lingjia Tang, Jason Mars
2015 Proceedings of the 42nd Annual International Symposium on Computer Architecture - ISCA '15  
We present and validate a methodology to study the impact of PCM on a datacenter, and evaluate two important opportunities for cost savings.  ...  However, this growth comes at the cost of an increasing thermal load that must be removed to prevent overheating and server failure.  ...  AMD, the AMD Arrow logo, and combinations thereof are trademarks of Advanced Micro Devices, Inc.  ... 
doi:10.1145/2749469.2749474 dblp:conf/isca/SkachAHLTTM15 fatcat:lco47taswfhtdktuabykigtnhy

Rubik

Harshad Kasture, Davide B. Bartolini, Nathan Beckmann, Daniel Sanchez
2015 Proceedings of the 48th International Symposium on Microarchitecture - MICRO-48  
Servers running these workloads are kept lightly loaded to meet these stringent latency targets. This low utilization wastes billions of dollars in energy and equipment annually.  ...  RubikColoc reduces datacenter power by up to 31% while using 41% fewer servers than a datacenter that segregates latency-critical and batch work, and achieves 100% core utilization.  ...  RUBIKCOLOC EVALUATION Experimental setup: To evaluate RubikColoc, we first consider a baseline datacenter that segregates batch and latency-critical apps.  ... 
doi:10.1145/2830772.2830797 dblp:conf/micro/KastureBBS15 fatcat:doivdbbc7rdkhfbixin5gpghbq

An analysis of power reduction in datacenters using heterogeneous chip multiprocessors

Vishal Gupta, Ripal Nathuji, Karsten Schwan
2011 Performance Evaluation Review  
In this paper, we present an opportunity analysis of heterogeneous chip multiprocessors in the context of datacenter environments where applications often have latency SLAs.  ...  Specifically, we define three use cases of heterogeneous processors for datacenter applications and adopt an analytical approach to quantify relative energy savings of using heterogeneous processors over  ...  The presence of SLAs in datacenters exhibits a similar opportunity. It allows applications to run on low-powered small cores even at a lower performance level as far as SLAs are met.  ... 
doi:10.1145/2160803.2160867 fatcat:hls2sxh6vncvpgfhlmjwbe7g34

On the Evaluation of the Total-Cost-of-Ownership Trade-offs in Edge vs Cloud deployments: A Wireless-Denial-of-Service Case Study

Panagiota Nikolaou, Yiannakis Sazeides, Alejandro Lampropoulos, Denis Guilhot, Andrea Bartoli, George Papadimitriou, Athanasios Chatzidimitriou, Dimitris Gizopoulos, Konstantinos Tovletoglou, Lev Mukhanov, Georgios Karakonstantis
2019 IEEE Transactions on Sustainable Computing  
Each Edge deployment site can contain one or even numerous servers, called, Cloudlets [17] and Micro DataCenters [18] .  ...  each hardware component (i.e. cores and memories) for enhancing the energy efficiency of micro-servers. ii) We evaluate the possible gains of an Edge deployment compared to a Cloud one, using a TCO model  ...  His research focuses on energy-efficient and error-resilient computing and storage architectures for embedded and highperformance applications.  ... 
doi:10.1109/tsusc.2019.2894018 fatcat:p4r2im3mzvcttabipj6w4ykouy

DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inference [article]

Udit Gupta, Samuel Hsia, Vikram Saraph, Xiaodong Wang, Brandon Reagen, Gu-Yeon Wei, Hsien-Hsin S. Lee, David Brooks, Carole-Jean Wu
2020 arXiv   pre-print
Finally, design, deployment, and evaluation in at-scale production datacenter shows over 30% latency reduction across a wide variety of recommendation models running on hundreds of machines.  ...  In this paper, we devise a novel end-to-end modeling infrastructure, DeepRecInfra, that adopts an algorithm and system co-design methodology to custom-design systems for recommendation use cases.  ...  Furthermore, recent work has opensourced benchmarks for studying the performance implication of at-scale execution of latency critical datacenter workloads and cloud micro-services [26] , [29] .  ... 
arXiv:2001.02772v1 fatcat:7dj2dy4rvnfynbvjov42jhzcc4

Automated Dynamic Resource Provisioning and Monitoring in Virtualized Large-Scale Datacenter

Sameera Abar, Pierre Lemarinier, Georgios K. Theodoropoulos, Gregory M.P. OHare
2014 2014 IEEE 28th International Conference on Advanced Information Networking and Applications  
On top of this, an Infrastructure-as-a-Service (IaaS) multi-tenancy usage model is constructed where users are given access to the systems capacity in the form of virtual servers and networks [2] .  ...  Currently, the multi-agent based applications incorporating the resource knowledge of the domain under consideration stay focused on the low intensity system design level.  ...  Runtime error information is matched with a predefined string or precondition. and an event is generated SimInvoke-Ag Linux-based Shell script NovaBoot.sh Clearing cache of inactive and delayed  ... 
doi:10.1109/aina.2014.117 dblp:conf/aina/AbarLTO14 fatcat:i6zfmjfmuzaq7fipcy3lan3aoa

VirtualPower

Ripal Nathuji, Karsten Schwan
2007 Proceedings of twenty-first ACM SIGOPS symposium on Operating systems principles - SOSP '07  
Experimental evaluations on modern multicore platforms highlight resulting improvements in online power management capabilities, including minimization of power consumption with little or no performance  ...  Finally, coordination of online methods for server consolidation with VPM management techniques in heterogeneous server systems is shown to provide up to 34% improvements in power consumption.  ...  ACKNOWLEDGMENTS We thank the anonymous reviewers and our shepherd, Mendel Rosenblum, for their feedback on improving this paper.  ... 
doi:10.1145/1294261.1294287 dblp:conf/sosp/NathujiS07 fatcat:jnhwtjragrh5fnu6l2ymmtweei
« Previous Showing results 1 — 15 out of 408 results