Block I logo
University of Illinois Urbana-Champaign
The Grainger College of Engineering
Parallel Programming Laboratory
Search
Menu
Search
Home
Research
Research
Libraries and Algorithms
Fine-grained Graph Algorithms
Parallel Algorithms and Applications
Reusable Libraries Software Engineering for Parallel Computing
Dense LU Factorization
Frameworks
AMR - Adaptive Mesh Refinement Framework
ParFUM - Parallel Framework for Unstructured Meshing
Parallel State Space Search Engine
Parallel Discrete Event Simulation
FEM - Finite Element Framework
Parallel Languages/Paradigms
Dagger and Structured Dagger
Charm++
AMPI - Adaptive Message Passing Interface
Automatic Communication Optimizations
MSA - Multiphase Shared Arrays Library in Charm++
Charj: Compiler Support for Productive Parallel Programming
OpenMP
Charm4py
Scientific Applications
ChaNGa - Computational Cosmology and N-Body Codes
NAMD - Molecular Dynamics
Parallel Operations Research
Other Applications
EpiSimdemics
OpenAtom - Ab Initio Molecular Dynamics
XPACC - The Center for Exascale Simulation of Plasma-Coupled Combustion
CEESD - The Center for Exascale-enabled Scramjet Design
Grid, Clusters, and Clouds
Faucets - Shared Computing Power
High Performance Computing in the Cloud
Runtime Systems and Tools
Dynamic Scripting for Parallel Applications
Parallel Debugging
BigNetSim - Parallel InterConnection Network Simulation
Dynamic Load Balancing
BlueWaters - Sustained Petascale Computing
Fault Tolerance Support
Parallel Performance Analysis, Visualization and Optimization
BigSim - Simulating PetaFLOPS Supercomputers
Heterogeneous Computing
Topology Aware Mapping
Power and Energy Efficiency
Interoperability
Performance-analysis-based Introspective Control System to Tune Applications and Runtime
Older Projects
Particles Framework
Multiblock Framework
Automatic Out of Core Execution Support
The Master-Slave Paradigm
Charisma++ - Orchestrating Parallel Objects
Standard Libraries for Parallel Programming
Cluster Computing at PPL
Papers
Software
Software
Charm++
Projections
Charm Debug
Charm License
Docs
People
People
Careers
Past Members
Home
Papers
Hierarchical Load Balancing for Charm++ Applications on Large Supercomputers
Hierarchical Load Balancing for Charm++ Applications on Large Supercomputers
Gengbin Zheng
Esteban Meneses
Abhinav Bhatele
Laxmikant Vasudeo Kale
International Workshop on Parallel Programming Models and Systems Software for High-End Computing at ICPP (P2S2) 2010
Pulication Type: Talk
Download:
pdf
ps
Abstract
This talk presents an automatic dynamic hierarchical load balancing method that overcomes the scalability challenges of centralized schemes and poor solutions of traditional distributed schemes.
Research Areas
Text Ref
BibTex