Filters








1,109 Hits in 7.0 sec

The Collection Virtual Machine: An Abstraction for Multi-Frontend Multi-Backend Data Analysis [article]

Ingo Müller
2020 arXiv   pre-print
In this paper, we propose the "Collection Virtual Machine" (or CVM) -- an extensible compiler framework designed to keep the specialization process of data analytics systems tractable.  ...  We have used CVM successfully to build specialized backends for platforms as diverse as multi-core CPUs, RDMA clusters, and serverless computing infrastructure in the cloud and expect similar results for  ...  Based on this hypothesis, we build the "Collection Virtual Machine" (CVM), a compiler framework for multi-frontend multi-backend data analysis.  ... 
arXiv:2004.01908v2 fatcat:7k2svf6sobghrexr7oaaaezgeu

The Collection Virtual Machine: An Abstraction for Multi-Frontend Multi-Backend Data Analysis

Ingo Müller, Renato Marroquín, Dimitrios Koutsoukos, Mike Wawrzoniak, Sabir Akhadov, Gustavo Alonso, Danica Porobic, Thomas Neumann
2020
Alonso 7 The Collection Virtual Machine CVM defines an IR language  IRs can evolve  IRs can be added  IRs can be combined  all IRs use the same framework | | Collection Types (Application-level) item  ...  all (many) frontends and backends  Streams as infinite collections?  ... 
doi:10.3929/ethz-b-000416413 fatcat:evyhwoy7a5dxbe2ow422udwkty

Abstractions for Practical Virtual Machine Replay

Anton Burtsev, David Johnson, Mike Hibler, Eric Eide, John Regehr
2016 Proceedings of the12th ACM SIGPLAN/SIGOPS International Conference on Virtual Execution Environments - VEE '16  
While implementing deterministic replay is hard, we argue that the main reason is the lack of general abstractions for understanding and addressing the significant engineering challenges involved in the  ...  We present a design blueprint-a set of abstractions, general principles, and low-level implementation details-for efficient deterministic replay in a modern hypervisor.  ...  Acknowledgments We thank the anonymous VEE '16 reviewers for their helpful comments. We performed our experiments on machines provided by the Utah Emulab testbed [47] .  ... 
doi:10.1145/2892242.2892257 dblp:conf/vee/BurtsevJHER16 fatcat:vszr2bmn7fg4doesvf65x36eny

A Fast and Transparent Communication Protocol for Co-Resident Virtual Machines

Yi Ren, Ling Liu, Xiaojian Liu, Jinzhu Kong, Huadong Dai, Qingbo Wu, Yuan Li
2012 Proceedings of the 8th IEEE International Conference on Collaborative Computing: Networking, Applications and Worksharing  
Network I/O workloads are dominating in most of the Cloud data centers today.  ...  In this paper, we argue that the design of inter-VM communication mechanism for co-resident VMs to replace TCP/IP based network virtualization should be completely transparent to applications running on  ...  ACKNOWLEDGMENT The first author's research is partially supported by grants from National Advanced Technology Research and  ... 
doi:10.4108/icst.collaboratecom.2012.250405 dblp:conf/colcom/RenLLKDWL12 fatcat:dg3dwz3yofdq5kpqrxfbyu7rai

PrivPy: Enabling Scalable and General Privacy-Preserving Machine Learning [article]

Yi Li, Yitao Duan, Yu Yu, Shuoyao Zhao, Wei Xu
2020 arXiv   pre-print
We introduce PrivPy, a practical privacy-preserving collaborative computation framework, especially optimized for machine learning tasks.  ...  PrivPy provides an easy-to-use and highly compatible Python programming front-end which supports high-level array operations and different secure computation engines to allow for security assumptions and  ...  The backend is linked to the frontend as a library on each server to reduce the overhead between the frontend and backend.  ... 
arXiv:1801.10117v7 fatcat:cxsy4nmoazfdzp3esn2etcn62m

Residency-Aware Virtual Machine Communication Optimization: Design Choices and Techniques

Yi Ren, Ling Liu, Qi Zhang, Qingbo Wu, Jie Wu, Jinzhu Kong, Jianbo Guan, Huadong Dai
2013 2013 IEEE Sixth International Conference on Cloud Computing  
One way to improve inter Virtual Machine (VM) communication efficiency is to support co-resident VM communication by using shared memory based approaches and to resort to the traditional TCP/IP for inter-VM  ...  We also give an analysis of future requirements in advanced features such as reliability, security and stability.  ...  The frontend and the backend exchange data by sharing memory pages, either in copying mode or in page flipping mode.  ... 
doi:10.1109/cloud.2013.117 dblp:conf/IEEEcloud/RenLZWWKGD13 fatcat:zq4vfjebqjeffdfojgg6g6bziy

Joeq: A virtual machine and compiler infrastructure

John Whaley
2005 Science of Computer Programming  
Joeq is released as open source software, and is being used as a framework by researchers all over the world on topics ranging from automatic distributed virtual machines to whole-program pointer analysis  ...  Joeq 1 is a virtual machine and compiler infrastructure designed to facilitate research in virtual machine technologies such as Just-In-Time and Ahead-Of-Time compilation, advanced garbage collection techniques  ...  I would also like to thank the students of the CS243 Winter 2002 class at Stanford for their feedback on the Joeq compiler infrastructure.  ... 
doi:10.1016/j.scico.2004.08.006 fatcat:ro5sz7szqvcqtgohjvrtfcvv4a

The hiphop virtual machine

Keith Adams, Jason Evans, Bertrand Maher, Guilherme Ottoni, Andrew Paroski, Brett Simmers, Edwin Smith, Owen Yamauchi
2014 Proceedings of the 2014 ACM International Conference on Object Oriented Programming Systems Languages & Applications - OOPSLA '14  
The HipHop Virtual Machine (HHVM) is a JIT compiler and runtime for PHP.  ...  Some types can be proven through static analysis, but limitations in the aheadof-time approach leave some types to be discovered at run time.  ...  Acknowledgments The HHVM project has been a huge engineering effort, which would not have been possible without the contributions and/or support of: Ali-Reza Adl- Tabatabai  ... 
doi:10.1145/2660193.2660199 dblp:conf/oopsla/AdamsEMOPSSY14 fatcat:ice77zutz5bz5drsgwfdpd3kfq

Multi-model Machine Learning Inference Serving with GPU Spatial Partitioning [article]

Seungbeom Choi, Sunho Lee, Yeonjae Kim, Jongse Park, Youngjin Kwon, Jaehyuk Huh
2021 arXiv   pre-print
To address the two requirements of ML inference servers, this paper proposes a new ML inference scheduling framework for multi-model ML inference servers.  ...  The scheduler assigns requests to virtual GPUs, called gpu-lets, with the most effective amount of resources.  ...  as a single virtual context, and 3) invoke the collection of kernels as if there is only a single context.  ... 
arXiv:2109.01611v1 fatcat:twpsj5ke4nazjgi4uio3cemini

LitSuggest: a web-based system for literature recommendation and curation using machine learning

Alexis Allot, Kyubum Lee, Qingyu Chen, Ling Luo, Zhiyong Lu
2021 Nucleic Acids Research  
However, it is challenging for a user to design optimal search queries using all the keywords related to a given topic.  ...  LitSuggest combines advanced machine learning techniques for suggesting relevant PubMed articles with high accuracy.  ...  ACKNOWLEDGEMENTS The authors would like to thank Alan James Bridge and his team for useful feedback and discussions on LitSuggest.  ... 
doi:10.1093/nar/gkab326 pmid:33950204 pmcid:PMC8262723 fatcat:7p6krmcj55d2noarufaco6cxti

The Design of the PROMIS Compiler [chapter]

Hideki Saito, Nicholas Stavrakos, Steven Carroll, Constantine Polychronopoulos, Alex Nicolau
1999 Lecture Notes in Computer Science  
PROMIS is a multilingual, parallelizing, and retargetable compiler with an integrated frontend and backend operating on a single unified/universal intermediate representation.  ...  The unified internal representation propagates hard to compute dependence information from the semantic rich frontend through the backend down to the code generator.  ...  Ackknowledgements The authors are grateful to Peter Grun, Ashok Halambi, and Nick Savoiu of University of California at Irvine for their work on the trailblazing part of the backend and pointer analysis  ... 
doi:10.1007/978-3-540-49051-7_15 fatcat:pwro2nysuvdr3mxz4ic6lob7ma

Joeq

John Whaley
2003 Proceedings of the 2003 workshop on Interpreters, Virtual Machines and Emulators - IVME '03  
Joeq is released as open source software, and is being used as a framework by researchers all over the world on topics ranging from automatic distributed virtual machines to whole-program pointer analysis  ...  Joeq 1 is a virtual machine and compiler infrastructure designed to facilitate research in virtual machine technologies such as Just-In-Time and Ahead-Of-Time compilation, advanced garbage collection techniques  ...  I would also like to thank the students of the CS243 Winter 2002 class at Stanford for their feedback on the Joeq compiler infrastructure.  ... 
doi:10.1145/858570.858577 fatcat:s5dx3mk3xnhkblls7zwf7yn5qy

Smile

Lei Cao, Jimeng Sun, Remco Chang, Brandon Westover, Samuel Madden, Michael Stonebraker, Wenbo Tao, Sungtae An, Jing Jin, Yizhou Yan, Xiaoyu Liu, Wendong Ge (+2 others)
2019 Proceedings of the VLDB Endowment  
However, the existing IIC classification techniques are shown to be not accurate and robust enough for clinical use because of the lack of high quality labels of EEG segments as training data.  ...  In this work, we propose Smile, an industrial scale system that provides an end-to-end solution to the IIC pattern classification problem.  ...  We are also grateful to Google Cloud for the GCP credits.  ... 
doi:10.14778/3352063.3352138 fatcat:yz2trncexzhp3c6455vx277p7i

FedApp: a Research Sandbox for Application Orchestration in Federated Clouds using OpenStack [article]

Johan Ruuskanen, Haorui Peng, Alfred Åkesson, Lars Larsson, Maria Kihl
2021 arXiv   pre-print
virtual clusters in an OpenStack environment with the possibility of adding realistic network characteristics between sites.  ...  Multi-cluster federation is envisioned to be the next-generation cloud infrastructure, where it will play a vital part in the realization of concepts such as edge and fog computing.  ...  Also, the work in this paper is partially supported by the SEC4FACTORY project, funded by the Swedish Foundation for Strategic Research (SSF), and the 5G-PERFECTA Celtic Next project funded by Sweden's  ... 
arXiv:2109.01480v1 fatcat:e7lij4xpk5f3voa25dp5qs4weq

JITA4DS: Disaggregated Execution of Data Science Pipelines Between the Edge and the Data Centre

Genoveva Vargas-Solar, Md Sahil Hassan, Ali Akoglu
2021 Journal of Web Engineering  
Vertical integration of these layers is needed for building a customizable Virtual Data Center (VDC) to meet the dynamically changing data science pipelines' requirements such as performance, availability  ...  Accordingly, the paper shows an experimental simulation devoted to run data science workloads and determine the best strategies for scheduling the allocation of resources implemented by JITA-4DS.  ...  Acknowledgements This material is based upon work supported by the National Science Foundation under Grant No. NSF CNS-1624668.  ... 
doi:10.13052/jwe1540-9589.2111 fatcat:b6gb4mbi7zbwviivxfju3tizvi
« Previous Showing results 1 — 15 out of 1,109 results