15,294 Hits in 2.2 sec

Spark Acceleration On Fpgas: A Use Case On Machine Learning In Pynq

Elias Koromilas, Ioannis Stamelos, Christoforos Kachris, Soudris. Dimitrios
2017 Zenodo  
the machine learning library's function with the function that invokes the hardware accelerator.  ...  The modifies functions for the execution of the function in the hardware accelerator, need to convert the RDDs to an array of variables, send the RDDs to the hardware accelerator and then convert back  ... 
doi:10.5281/zenodo.801507 fatcat:d4cav7b3dre3vkrpvhieqde2au

FaaM: FPGA-as-a-Microservice - A Case Study for Data Compression

David Ojika, Ann Gordon-Ross, Herman Lam, Bhavesh Patel, A. Forti, L. Betev, M. Litmaath, O. Smirnova, P. Hristov
2019 EPJ Web of Conferences  
To address these domains' processing needs, recent research has focused on using FPGAs to accelerate workloads, ranging from analytics and machine learning to databases and network function virtualization  ...  to straightforward implementation with minimal to no communication overhead despite the hardware abstraction.  ...  FPGA Accelerator Anknoledgement Scheduler To schedule cloud users' jobs, we focus on a task scheduler that is local to each Worker Container.  ... 
doi:10.1051/epjconf/201921407029 fatcat:3z6eji6mj5azfgixp5uowjdu3i

Enabling FPGAs in the cloud

Fei Chen, Yi Shan, Yu Zhang, Yu Wang, Hubertus Franke, Xiaotao Chang, Kun Wang
2014 Proceedings of the 11th ACM Conference on Computing Frontiers - CF '14  
The prototype enables isolation between multiple processes in multiple VMs, precise quantitative acceleration resource allocation, and priority-based workload scheduling.  ...  As more and more workloads are being deployed in the cloud, it is appropriate to consider how to make FPGAs and their capabilities available in the cloud.  ...  These projects focused on accelerator scheduling in reconfigurable hardware based on non-virtualized environments.  ... 
doi:10.1145/2597917.2597929 dblp:conf/cf/ChenSZWFCW14 fatcat:l5lhpbl4mfa4hgohs5swcb4fli

D2.3: System Architecture

Christoforos Kachris, Angelos Bilas, Nikos Chrysos, Hans Vandierendonck
2017 Zenodo  
VINEYARD's goal is to both develop energy efficient hardware-accelerated servers and to develop the required framework for the seamless utilization of these servers in the programming frameworks that are  ...  are developed for the efficient integration of the hardware platforms.  ...  These hardware accelerators are stored in an IP repository (VineStore) that interface with the VINEYARD resource manager and scheduler.  ... 
doi:10.5281/zenodo.898155 fatcat:gjqv2gfngfev3bmmrlxycgzv64

The VINEYARD Approach: Versatile, Integrated, Accelerator-Based, Heterogeneous Data Centres [chapter]

Christoforos Kachris, Dimitrios Soudris, Georgi Gaydadjiev, Huy-Nam Nguyen, Dimitrios S. Nikolopoulos, Angelos Bilas, Neil Morgan, Christos Strydis, Christos Tsalidis, John Balafas, Ricardo Jimenez-Peris, Alexandre Almeida
2016 Lecture Notes in Computer Science  
VINEYARD aims to develop an integrated platform for energy-e cient data centres based on new servers with novel, coarse-grain and fine-grain, programmable hardware accelerators.  ...  This programming framework will, further, allow the hardware accelerators to be swapped in and out of the heterogeneous infrastructure so as to o↵er high flexibility and energy e ciency.  ...  These hardware accelerators can be hosted in a repository that will interface with the run-time scheduler.  ... 
doi:10.1007/978-3-319-30481-6_1 fatcat:4yvigra2w5aa7dakazch6x3qni

Online scheduling for FPGA computation in the Cloud

Guohao Dai, Yi Shan, Fei Chen, Yu Wang, Kun Wang, Huazhong Yang
2014 2014 International Conference on Field-Programmable Technology (FPT)  
The results show that our FPGA accelerated cloud system is 1.386 times faster than using the previous algorithm.  ...  In this paper, we propose a benefit-based scheduling metric to evaluate the task assignment. Based on the metric, we accelerate task execution according to our benefit-based scheduling algorithms.  ...  In the hardware layer, an accelerator pool (AP) abstraction is proposed to use FPGAs in the cloud.  ... 
doi:10.1109/fpt.2014.7082811 dblp:conf/fpt/DaiSCWWY14 fatcat:up7nmqidcrculhx4gkdvj5og6i

At the Edge of a Seamless Cloud Experience [article]

Samuel Rac, Mats Brorsson
2021 arXiv   pre-print
The traditional cloud computing paradigm can not meet this requirement, legitimizing the need for a new paradigm.  ...  In this paper, we discuss the main challenges to be met in edge computing and solutions to achieve a seamless cloud experience.  ...  In addition, hardware accelerators have to be taken into account.  ... 
arXiv:2111.06157v1 fatcat:khos7gl7l5drpgv5x4mgdrfjum

Heterogeneous Resource Management and Orchestration in Cloud Environments [chapter]

Dapeng Dong, Huanhuan Xiong, Gabriel G. Castañé, Paul Stack, John P. Morrison
2018 Communications in Computer and Information Science  
The accelerated uptake of heterogeneous resources is exacerbating these challenges, which no longer can be efficiently addressed in an ad-hoc manner.  ...  The addition of heterogeneous resources to conventional homogeneous cloud environments has enabled clouds to embrace a wide variety of new applications that heretofore were traditionally confined to specialized  ...  these hardware resources and accelerators.  ... 
doi:10.1007/978-3-319-94959-8_4 fatcat:7eoeuexrt5amdd2eirvhlfhsai

Heterogeneous Cloud Computing: The Way Forward

Stephen P. Crago, John Paul Walters
2015 Computer  
A major appeal of cloud computing is that it abstracts hardware architecture from both end users and programmers.  ...  Extending homogeneous cloud flexibility to heterogeneous IaaS deployment requires further research in several areas: › optimal tradeoffs in virtualization performance and functionality (security vis à  ... 
doi:10.1109/mc.2015.14 fatcat:chu656zg3ffcdcgum34vas3b2q

Embedded Deep Learning Prototyping Approach for Cyber-Physical Systems: Smart LIDAR Case Study

Quentin Cabanes, Benaoumeur Senouci, Amar Ramdane-Cherif
2021 Journal of Sensor and Actuator Networks  
The input of our NNP is a voxel grid hardware computed from 3D point cloud. Finally, the results show that our NNP is able to process Dense Neural Network (DNN) architecture without bias.  ...  These CPSs deal with data analysis, which need powerful algorithms combined with robust hardware architectures. On one hand, Deep Learning (DL) is proposed as the main solution algorithm.  ...  This NNP could be considered a fully functional Intellectual Property (IP) to be integrated alongside the other IPs from the hardware-accelerated embedded processing.  ... 
doi:10.3390/jsan10010018 fatcat:efdfu3bkmvha5ay6ny6ljcm2c4

A Survey of NFV Network Acceleration from ETSI Perspective

Yong-Xuan Huang, Jerry Chou
2022 Electronics  
We expect that NFV will increasingly rely on cloud services in the future. Since cloud services do not offer a choice of hardware, our acceleration method will be primarily software-based.  ...  Network function virtualization (NFV) enables network operators to save costs and flexibility by replacing dedicated hardware with software network functions running on commodity servers.  ...  Schedule/Predit Table 8 . 8 An ETSI framework perspective on acceleration strategies that can be used in cloud infrastructure.  ... 
doi:10.3390/electronics11091457 fatcat:3cpfsn7rirhnjglhrt2rah24im

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  
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  ...  The isolation is enabled by introducing a two-level instruction dispatch module and a multi-core based hardware resources pool.  ...  It is important to mention that the scheduling algorithm is mainly used in the private cloud scenarios.  ... 
doi:10.1145/3480170 fatcat:kgrhiohisvcdxm635l3wykgdri

Benefits and Challenges of Cloud Technologies for 5G Architecture

Dario Sabella, Peter Rost, Albert Banchs, Valentin Savin, Marco Consonni, Marco Di Girolamo, Massinissa Lalam, Andreas Maeder, Ignacio Berberana
2015 2015 IEEE 81st Vehicular Technology Conference (VTC Spring)  
First, we provide a comprehensive overview of implementation aspects and how different hardware options impact the implementation of RAN functionality.  ...  This paper focuses on the practical implementation of a Cloud-RAN architecture in the context of future 5G systems, with particular emphasis on different aspects of the functional split between the cloud  ...  Furthermore, hybrid approaches are possible where a software implementation on GPPs is complemented with hardware accelerators.  ... 
doi:10.1109/vtcspring.2015.7145716 dblp:conf/vtc/SabellaRBSCGLMB15 fatcat:sqlaib3leza3fowqpu7tplsaoe

Enabling FPGA-as-a-Service in the Cloud with hCODE Platform

Qian ZHAO, Motoki AMAGASAKI, Masahiro IIDA, Morihiro KUGA, Toshinori SUEYOSHI
2018 IEICE transactions on information and systems  
Efficient FPGA virtualization and accelerator scheduling techniques are proposed to deploy accelerators on the FPGA-enabled cluster easily.  ...  With the proposed hCODE, hardware designers and accelerator users can be organized on one platform to efficiently build open-hardware ecosystem.  ...  We are also going to evaluate future work on public FPGA-enabled clouds.  ... 
doi:10.1587/transinf.2017rcp0004 fatcat:n7humdedpjhfxc7jtw6rkxkh44

On the Optimization of Self-Organization and Self-Management Hardware Resource Allocation for Heterogeneous Clouds

Konstantinos M. Giannoutakis, Christos K. Filelis-Papadopoulos, George A. Gravvanis, Dimitrios Tzovaras
2021 Computers  
There is a tendency, during the last years, to migrate from the traditional homogeneous clouds and centralized provisioning of resources to heterogeneous clouds with specialized hardware governed in a  ...  In this work, an optimized Suitability Index and assessment function are proposed, along with their theoretical analysis, for improving the computational efficiency, energy consumption, service delivery  ...  By also adopting heterogeneity and specialized hardware (accelerators) for supporting computations, the efficient scheduling of resources is very important.  ... 
doi:10.3390/computers10110147 fatcat:uor6wxxgovawnd7zdrcyvxho2q
« Previous Showing results 1 — 15 out of 15,294 results