12 Hits in 3.5 sec

Rolling Bearing Fault Diagnosis Method Based on Parallel QPSO-BPNN under Spark-GPU Platform

Lanjun Wan, Hongyang Li, Gen Zhang, Changyun Li, Junfeng Man, Mansheng Xiao
2021 IEEE Access  
Compared with MapReduce, Spark introduces resilient distributed data set (RDD) and implements an efficient directed acyclic graph execution engine, it has a faster processing speed, and thus it is more  ...  PROCESS OF ROLLING BEARING FAULT DIAGNOSIS BASED ON PARALLEL QPSO-BPNN The overall process of rolling bearing fault diagnosis based on parallel QPSO-BPNN under Spark-GPU platform is shown in Fig. 2 ,  ... 
doi:10.1109/access.2021.3072596 fatcat:6vdrxp7z7vbu7i3qqjktjsk3em

Collaborative Filtering Recommendation Using Nonnegative Matrix Factorization in GPU-Accelerated Spark Platform

Bing Tang, Linyao Kang, Li Zhang, Feiyan Guo, Haiwu He, Shah Nazir
2021 Scientific Programming  
The new GPU-accelerated NMF on Spark platform is evaluated in a 4-node Spark heterogeneous cluster using Google Compute Engine by configuring each node a NVIDIA K80 CUDA device, and experimental results  ...  To solve this problem, this paper proposes a parallel algorithm based on GPU for NMF in Spark platform, which makes full use of the advantages of in-memory computation mode and GPU acceleration.  ...  which is proposed in this paper and developed on Spark-GPU fusion platform.  ... 
doi:10.1155/2021/8841133 fatcat:6tf7qm7zwzce3ebcas7r6dvum4

When Spark Meets FPGAs: A Case Study for Next-Generation DNA Sequencing Acceleration

Yu-Ting Chen, Jason Cong, Zhenman Fang, Jie Lei, Peng Wei
2016 2016 IEEE 24th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM)  
In this paper we aim to answer one key question: how can we efficiently integrate FPGAs into stateof-the-art big-data computing frameworks like Apache Spark?  ...  solutions including batch processing and the FPGA-as-a-Service framework to address them.  ...  Acknowledgments This work was supported in part by C-FAR, one of the six SRC STARnet Centers, sponsored by MARCO and DARPA, and by NSF/Intel Innovation Transition Grant awarded to the Center for Domain-Specific  ... 
doi:10.1109/fccm.2016.18 dblp:conf/fccm/ChenCFLW16 fatcat:zs3llx2cqrbyfi6dpqyyi5d3ne

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)  
-Taverne project Otherwise as indicated in  ...  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.  ...  In [62] , authors present an RDMA-based Spark on InfiniBand-enabled clusters in which the performance of SQL and graph processing workloads are improved by 32% and 46% respectively.  ... 
doi:10.1109/ispdc.2019.00011 dblp:conf/ispdc/RellermeyerKGP19 fatcat:kczllhax3rgkbeuhmblp6wsowq

A Survey on Spark Ecosystem for Big Data Processing [article]

Shanjiang Tang, Bingsheng He, Ce Yu, Yusen Li, Kun Li
2018 arXiv   pre-print
Finally, we make a discussion on the open issues and challenges for large-scale in-memory data processing with Spark.  ...  Arguably, Spark is state of the art in large-scale data computing systems nowadays, due to its good properties including generality, fault tolerance, high performance of in-memory data processing, and  ...  Its implementation is publicly available ( 4).Columnar RDD.  ... 
arXiv:1811.08834v1 fatcat:6fxvg6me7rayzm4suoabyg7fii

TCUDB: Accelerating Database with Tensor Processors [article]

Yu-Ching Hu and Yuliang Li and Hung-Wei Tseng
2021 arXiv   pre-print
Matrix multiplication was considered inefficient in the past; however, this strategy has remained largely unexplored in conventional GPU-based databases, which primarily rely on vector or scalar processing  ...  We present TCUDB, a TCU-accelerated query engine processing a set of query operators including natural joins and group-by aggregates as matrix operators within TCUs.  ...  Spark-GPU: evaluation and benchmarking. 237–252. An accelerated in-memory data processing engine on clusters.  ... 
arXiv:2112.07552v1 fatcat:y6furfuc7nh5jml3cew7mqngry

Pushing Big Data into Accelerators: Can the JVM Saturate Our Hardware? [chapter]

Johan Peltenburg, Ahmad Hesam, Zaid Al-Ars
2017 Lecture Notes in Computer Science  
One major challenge in combining JVM-based big data frameworks with accelerators is transferring data from objects that reside in JVM managed memory to the accelerator.  ...  Advancements in the field of big data have led into an increasing interest in accelerator-based computing as a solution for computationally intensive problems.  ...  This work was supported by the European Commission in the context of the ARTEMIS project ALMARVI (project #621439).  ... 
doi:10.1007/978-3-319-67630-2_18 fatcat:3nomjk76o5hxjkfzhzwynznidq

Apache Spark Accelerated Deep Learning Inference for Large Scale Satellite Image Analytics [article]

Dalton Lunga, Jonathan Gerrand, Hsiuhan Lexie Yang, Christopher Layton, Robert Stewart
2019 arXiv   pre-print
RESFlow takes advantage of both a unified analytics engine for large-scale data processing and the availability of modern computing hardware to harness the acceleration of deep learning inference on expansive  ...  As a consequence, this rapid advancement poses new computational and data processing challenges.  ...  The models were trained on a regular DGX-1 machine (top) and on a SPARK GPU cluster (bottom). Table I : I Sources and number of labelled data for training, validation and testing evaluation.  ... 
arXiv:1908.04383v1 fatcat:z33cgcot6fhwxpctb2b43phmqa

Apache Spark Accelerated Deep Learning Inference for Large Scale Satellite Image Analytics

Dalton Lunga, Jonathan Gerrand, Lexie Yang, Christopher Layton, Robert Stewart
2020 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
The workflow is demonstrated with NVIDIA DGX accelerated platforms and offers appreciable compute speed-ups for deep learning inference on pixel labeling workloads; processing 21 028 TB of imagery data  ...  We showcase its deployment in both computationally and data-intensive workloads for pixel-level labeling tasks.  ...  Fig. 7 . 7 F1 scores on validation set for the tested CNNs. The models were trained on a regular DGX-1 machine (top) and on a SPARK GPU cluster (bottom).  ... 
doi:10.1109/jstars.2019.2959707 fatcat:xaarxfgvwvgsxjzxpebhajxfhe

D5.1 Operator Cost Estimation and Workflow Optimisation Technology V1

Project Consortium Members
2020 Zenodo  
potentially geo-dispersed computer clusters each hosting one or more Big Data platforms.  ...  Moreover, WP5 interacts with the Synopses Data Engine C [...]  ...  Spark-GPU [YSH+16] is another solution that enables the execution of applications in both CPU and GPU, while extending/improving Spark to avoid overheads with memory copies and scheduling, providing  ... 
doi:10.5281/zenodo.4034108 fatcat:t22h4qqgjfbsporpl4zkf5c2qm

Go Meta! A Case for Generative Programming and DSLs in Performance Critical Systems 1 The Cost of Performance under Creative Commons License CC-BY 1st Summit on Advances in Programming Languages (SNAPL'15)

Tiark Rompf, Kevin Brown, Hyoukjoong Lee, Arvind Sujeeth, Manohar Jonnalagedda, Nada Amin, Georg Ofenbeck, Alen Stojanov, Yannis Klonatos, Mohammad Dashti, Christoph Koch, Markus Püschel (+20 others)
This design fits naturally with a distinction into control and data paths, which already exists in many systems.  ...  We argue that this needs to change, and that generative programming is an effective avenue to enable the use of high-level languages and programming techniques in many such circumstances.  ...  The work presented in this paper has profited from numerous interac-  ... 

Exploiting BSP Abstractions for Compiler Based Optimizations of GPU Applications on multi-GPU Systems

Alexander Matz
Graphics Processing Units (GPUs) are accelerators for computers and provide massive amounts of computational power and bandwidth for amenable applications.  ...  Next, an application model based on Z-polyhedra is derived that formalizes the distribution of work across multiple GPUs and allows the identification of data dependencies.  ...  His guidance throughout this process has helped my professional and personal growth tremendously and made this work possible.  ... 
doi:10.11588/heidok.00029213 fatcat:3spb4qu7h5bozojzaoebwo2brm