Special Issue of Sustainable Computing: Informatics and Systems
(SUSCOM) on Energy-Aware Resource Management and Scheduling (EARMS)
Scope: The growing scale of High Performance Computing (HPC) systems
and data centers has made issues related to power consumption, air
conditioning, and cooling infrastructures critical concerns in
terms of efficiency, operational cost, reliability, energy
conservation, and environmental impact. High-end HPC systems today
consume several megawatts of power, enough to power small towns,
and are, in fact, soon approaching the limits of the power
available to them. Furthermore, the costs of powering these HPC
systems runs into millions of dollars per year, and are increasing as
these systems target sustained petascale and plan for exascale. Adding
to the concerns due to power and cooling requirements and associated
costs, empirical data show that every 10 degree Celsius increase in
temperature results in a doubling of the system failure rate, which
reduces the reliability of these expensive systems. At the same time,
the increasing proliferation of virtualization technologies and the
consolidation of computing platforms for data- and compute-intensive
applications are providing new opportunities to use advanced
scheduling and resource management techniques for higher utilization
and energy savings. As a result, resource management policies that
consider the tradeoffs between energy usage and performance,
throughput, and other QoS requirements, have become important research
challenges that must be addressed.
In this special issue, we seek original work focused on addressing new
research and development challenges, consisting of developing new
scheduling techniques and advanced resource management solutions
in green computing for HPC systems.
Specific topics include, but not limited to, the following:
Energy-aware task scheduling heuristics
Energy-aware resource management infrastructures
Thermal-aware task scheduling heuristics
Thermal-aware resource management infrastructures
Energy and thermal-aware scheduling for fault-tolerance
Enhanced performance, energy, and thermal models for energy-aware
schedulers
Resource management in power/energy constrained systems
Energy efficiency and virtualization
Energy, performance, quality of service, and other resource tradeoffs
Energy-aware scheduling for heterogeneous systems
Energy-aware scheduling for parallel and distributed systems
Energy-efficient scheduling of hardware accelerators (FPGAs, GPUs,
etc.)
Energy-aware scheduling for different computing paradigms: cluster,
grid, cloud, and datacenters
Submission Details:
General information for submitting papers to SUSCOM can be found at
http://ees.elsevier.com/
link). Submissions to this Special Issue (SI) should be made using
Elsevier's editorial system at the journal website
(http://ees.elsevier.com/
make sure to select the ?SI: En. Aware Sched.? option for the type of
the paper during the submission process. All submissions must be
original and may not be under review. A submission based on one or more
papers that appeared elsewhere has to include major value-added
extensions over what appeared previously (at least 30% new conceptual
material). Authors are requested to attach to the submitted paper such
earlier articles and a summary document explaining the enhancements made
in the journal version. All submitted papers will be peer reviewed using
the normal standards of SUSCOM. By submitting a paper to this special
issue, the authors agree to review at least one paper within the time
frame of the SI.
Important Dates:
Title and keywords due date (by email to EARMS@colostate.edu): January
15, 2014
Manuscript due date (on http://ees.elsevier.com/
2014
First decision notification: May 01, 2014
Revised manuscript due date: June 01, 2014
Final decision notification: July 01, 2014
Tentative publication schedule: October 2014
Special Issue Guest Editors:
Anne Benoit, ENS-Lyon, France
Ryan Friese, Colorado State University, U.S.A
H.J. Siegel, Colorado State University, U.S.A
Questions may be sent to EARMS@colostate.edu
(SUSCOM) on Energy-Aware Resource Management and Scheduling (EARMS)
Scope: The growing scale of High Performance Computing (HPC) systems
and data centers has made issues related to power consumption, air
conditioning, and cooling infrastructures critical concerns in
terms of efficiency, operational cost, reliability, energy
conservation, and environmental impact. High-end HPC systems today
consume several megawatts of power, enough to power small towns,
and are, in fact, soon approaching the limits of the power
available to them. Furthermore, the costs of powering these HPC
systems runs into millions of dollars per year, and are increasing as
these systems target sustained petascale and plan for exascale. Adding
to the concerns due to power and cooling requirements and associated
costs, empirical data show that every 10 degree Celsius increase in
temperature results in a doubling of the system failure rate, which
reduces the reliability of these expensive systems. At the same time,
the increasing proliferation of virtualization technologies and the
consolidation of computing platforms for data- and compute-intensive
applications are providing new opportunities to use advanced
scheduling and resource management techniques for higher utilization
and energy savings. As a result, resource management policies that
consider the tradeoffs between energy usage and performance,
throughput, and other QoS requirements, have become important research
challenges that must be addressed.
In this special issue, we seek original work focused on addressing new
research and development challenges, consisting of developing new
scheduling techniques and advanced resource management solutions
in green computing for HPC systems.
Specific topics include, but not limited to, the following:
Energy-aware task scheduling heuristics
Energy-aware resource management infrastructures
Thermal-aware task scheduling heuristics
Thermal-aware resource management infrastructures
Energy and thermal-aware scheduling for fault-tolerance
Enhanced performance, energy, and thermal models for energy-aware
schedulers
Resource management in power/energy constrained systems
Energy efficiency and virtualization
Energy, performance, quality of service, and other resource tradeoffs
Energy-aware scheduling for heterogeneous systems
Energy-aware scheduling for parallel and distributed systems
Energy-efficient scheduling of hardware accelerators (FPGAs, GPUs,
etc.)
Energy-aware scheduling for different computing paradigms: cluster,
grid, cloud, and datacenters
Submission Details:
General information for submitting papers to SUSCOM can be found at
http://ees.elsevier.com/
link). Submissions to this Special Issue (SI) should be made using
Elsevier's editorial system at the journal website
(http://ees.elsevier.com/
make sure to select the ?SI: En. Aware Sched.? option for the type of
the paper during the submission process. All submissions must be
original and may not be under review. A submission based on one or more
papers that appeared elsewhere has to include major value-added
extensions over what appeared previously (at least 30% new conceptual
material). Authors are requested to attach to the submitted paper such
earlier articles and a summary document explaining the enhancements made
in the journal version. All submitted papers will be peer reviewed using
the normal standards of SUSCOM. By submitting a paper to this special
issue, the authors agree to review at least one paper within the time
frame of the SI.
Important Dates:
Title and keywords due date (by email to EARMS@colostate.edu): January
15, 2014
Manuscript due date (on http://ees.elsevier.com/
2014
First decision notification: May 01, 2014
Revised manuscript due date: June 01, 2014
Final decision notification: July 01, 2014
Tentative publication schedule: October 2014
Special Issue Guest Editors:
Anne Benoit, ENS-Lyon, France
Ryan Friese, Colorado State University, U.S.A
H.J. Siegel, Colorado State University, U.S.A
Questions may be sent to EARMS@colostate.edu
No comments:
Post a Comment