Session01-VIRT

The Second Special Session on Virtualization in High Performance Computing and Simulation

(VIRT 2019)

CALL FOR PAPERS & PARTICIPATION

As part of

The 17th International Conference on High Performance Computing & Simulation (HPCS 2019)

http://hpcs2019.cisedu.info/ or http://conf.cisedu.info/rp/hpcs19

15 - 19 July 2019

Dublin, Ireland

Paper Submission Deadline: 15 April 2019 - Extended

Submissions could be for full papers, short papers, poster papers, or posters

PURPOSE and SCOPE

Virtualization has become the foundation of cloud computing. The scale of cloud computing data centers is similar to the supercomputers used in High Performance Computing. Both involve millions of CPU cores, tens of thousands of accelerators like GPUs, FPGAs connected by high speed interconnects and different connection topologies. Studies have shown that the GPU utilization in HPC is often at 50%. Virtualization can be the key to increasing resource utilization and managing massive infrastructures efficiently.

GPU vendors have introduced virtualized versions of GPUs. FPGAs are being deployed in virtual infrastructure. These and other developments in the field, along with future exascale systems, will provide increasing degree of virtualization within the systems.

The goal of this session is to be an opportunity to discuss and exchange research on the different virtualization technologies and how they can be efficiently applied in High Performance Computing. Theoretical research, engineering solutions dealing with practical tradeoffs, and complex system implementation papers are welcomed.

The VIRT topics include (but are not limited to) the following:

  • Virtualization in HPC

o Energy efficient solutions for HPC applications in the cloud environment

o High-availability based on system-level virtualization mechanisms for HPC

o Fault tolerance mechanisms based on system-level virtualization mechanisms for HPC

o Performance and evaluation of diverse HPC workloads in virtualization environment

o Security Isolation, Resource Isolation and Data Isolation in Virtualized HPC

o Flexibility and Ease of system administration and management for virtualized environments for HPC

o Running diverse workloads

o Running heterogeneous Operating Systems

o Live Migration and Suspend/Resume of VMs to improve resource utilization and avoid downtime

o Scheduling of resources and Workloads

o Software Defined Data Centers (SDDC)

o Hypervisors and other virtualization solutions tailored for HPC systems

o Big Data and Analytics in Virtualized Environments

o Enterprise High Performance Computing in Virtualized Environments

  • Accelerators in Virtualized HPC Environments:

o Virtualized GPUs

§ Scheduling of Virtual GPUs

o Using FPGAs in Virtualization Stack

§ Dynamic Reconfiguration of FPGAs

  • IO Virtualization

o Network Virtualization

o SR-IOV

o Software Defined Networking (SDN)

o Storage Virtualization

o RDMA and vRDMA

  • Containers in HPC

o Isolation of Multiple User-Spaces

o Running containers in Virtual Machines

  • Tools

o Virtualization solutions for dealing with heterogeneity in HPC environments

o Compilers for heterogenous architectures involving GPUS, FPGAs, multi-core CPUs

o Operating system support for virtualization in HPC systems

o Debugging and/or profiling in virtual environments

  • Algorithms

Uday Kurkure

VMware, Inc.

3401 Hillview Avenue

Palo Alto, CA 94304, USA

Phone: +1 650-427-1179

Fax: +1 650-

Email: uday@alumni.stanford.edu

Hari Sivaraman

VMware, Inc.

3401 Hillview Avenue

Palo Alto, CA 94304, USA

Phone: +1 650-427-3681

Fax: +1 650-

Email: hsivaraman@vmware.com

Lan Vu

VMware, Inc.

3401 Hillview Avenue

Palo Alto, CA 94304, USA

Phone: +1 650-427-1327

Fax: +1 650-

Email: lanv@vmware.com

International Program Committee*:

All submitted papers will be rigorously reviewed by the special session technical program committee members following similar criteria used in HPCS 2019 and will be published as part of the HPCS 2019 Proceedings.

  • Cristina Boeres, Fluminense Federal University, Brazil

  • Isaac Gelado, Nvidia Corp., Santa Clara, California, USA

  • Kyle Hale, Illinois Institute of Technology, Illinois, USA

  • Brian Kocoloski, Washington University in St. Louis, Missouri, USA

  • Xiaoyi Lu, The Ohio State University - Columbus, Ohio, USA

  • Carlos Reano, Queen's University Belfast, U.K.

  • Federico Silla, Universitat Politecnica de Valencia, Spain

  • Alex Sim, Lawrence Berkeley National Laboratory, California, USA

  • Giang Son Tran, University of Science and Technology of Hanoi, Hanoi, Vietnam

  • Blesson Varghese, Queen's University Belfast, U.K.

  • Andrew Younge, Sandia National Laboratories, Albuquerque, New Mexico, USA

  • Dong Ping Zhang, AMD Inc., San Jose, California, USA

  • Jie Zhang, Amazon Inc., Washington, USA

  • Na Zhang, VMware Inc., Boston, Massachusetts, USA

  • Yongli Zhao, Beijing University of Posts and Telecommunications, Beijing, China

(* Committee formation is pending and will be finalized shortly.)

For information or questions about Conference's paper submission, tutorials, posters, workshops, special sessions, exhibits, demos, panels and forums organization, doctoral colloquium, and any other information about the conference location, registration, paper formatting, etc., please consult the Conference’s web site at URL: http://hpcs2019.cisedu.info/ or http://conf.cisedu.info/rp/hpcs19 or contact one of the Conference's organizers.

o Distributed Resource Scheduling

o Distributed Computing

  • AI, Machine Learning and Deep Learning

  • o Virtualization of Deep Learning Workloads

  • o Deep Learning in HPC Simulations

  • o Resource Allocation Using Machine Learning

  • o Reinforcement Learning for Resource Allocation

  • o Machine Learning to build tools for measurement & calibration

  • Virtualization in Simulation

INSTRUCTIONS FOR PAPER SUBMISSIONS

You are invited to submit original and unpublished research works on above and other topics related to virtualization in high performance computing and simulation. Submitted papers must not have been published or simultaneously submitted elsewhere until it appears in HPCS proceedings, in the case of acceptance, or notified otherwise. Submission can be for

- Regular papers, please submit a PDF copy of your full manuscript, not to exceed 8 double-column formatted pages per template, and include up to 6 keywords and an abstract of no more than 400 words. Additional pages will be charged additional fee. Submission should include a cover page with authors' names, affiliation addresses, fax numbers, phone numbers, and all authors email addresses. Please, indicate clearly the corresponding author(s) although all authors are equally responsible for the manuscript.

- Short papers (up to 4 pages), please submit a PDF copy of your full manuscript, not to exceed 4 double-column formatted pages per template, and include up to 6 keywords and an abstract of no more than 400 words. Additional pages will be charged additional fee. Submission should include a cover page with authors' names, affiliation addresses, fax numbers, phone numbers, and all authors email addresses. Please, indicate clearly the corresponding author(s) although all authors are equally responsible for the manuscript.

- Poster papers and Posters (please refer to http://hpcs2019.cisedu.info/1-call-for-papers-and-participation/call-for-posters for posters submission details) will also be considered.

Please specify the type of submission you have. Please include page numbers on all preliminary submissions to make it easier for reviewers to provide helpful comments.

Submit a PDF copy of your full manuscript to the special session paper submission site at https://cmt3.research.microsoft.com/VIRT2019 . Acknowledgement will be sent within 48 hours of submission.

Conference Policies

Only PDF files will be accepted, uploaded to the submission link above. Each paper will receive a minimum of three reviews. Papers will be selected based on their originality, relevance, significance, technical clarity and soundness, presentation, language, and references. Submission implies the willingness of at least one of the authors to register and present the paper, if accepted. At least one of the authors of each accepted paper will have to register and attend the HPCS 2019 conference to present the paper at the special session as scheduled. By submitting the paper to the HPCS conference, all authors agree to abide by all HPCS conference paper submission, publication and presentation policies as well as following ethical and professional codes of conduct, including those of the professional co-sponsoring organizations. Contents of manuscripts submitted to the tracks program committees shall be regarded as privileged as well and handled in the same manner and standards. For more information, please refer to the Authors Info and Registration Info pages.

Proceedings

Accepted papers will be published in the Conference proceedings. Instructions for final manuscript format and requirements will be posted on the HPCS 2019 Conference web site. It is our intent to have the proceedings formally published in hard and soft copies and be available at the time of the conference. The proceedings is projected to be included in the IEEE or ACM Digital Library and indexed in all major indexing services accordingly.

SPECIAL ISSUE

Plans are underway to have the best papers, in extended version, selected for possible publication in a reputable journal as special issue. Detailed information will soon be announced and will be made available on the conference website.

If you have any questions about paper submission or the special session, please contact the special session organizers.

IMPORTANT DATES

Paper Submissions: ------------------------------------------- 15 April 2019 - Extended

Acceptance Notification: -------------------------------------- 28 April 2019

Camera Ready Papers and Registration Due by: ----------- 10 May 2019

Conference Dates: -------------------------------------------- 15–19 July 2019

SPECIAL SESSION ORGANIZERS