ROMA - 2013 - Annual activity report

ROMA

ROMA - 2013

Team Roma

Members

Overall Objectives

Introduction
Highlights of the Year

Research Program

Algorithms for probabilistic environments
Platform-aware scheduling strategies
High-performance computing and linear algebra

Application Domains

Application of sparse direct solvers

Software and Platforms

MUMPS

New Results

Scheduling tree-shaped task graphs to minimize memory and makespan
Model and complexity results for tree traversals on hybrid platforms
On the combination of silent error detection and checkpointing
Checkpointing algorithms and fault prediction
Mapping applications on volatile resources
Using group replication for resilience on exascale systems
Unified model for assessing checkpointing protocols at extreme-scale
Revisiting the double checkpointing algorithm
Multi-criteria checkpointing strategies: Optimizing response-time versus resource utilization
Optimal checkpointing period: Time vs. energy
Energy-aware checkpointing of divisible tasks with soft or hard deadlines
Assessing the performance of energy-aware mappings
Computing the throughput of probabilistic and replicated streaming applications
Reliability and performance optimization of pipelined real-time systems
Scheduling linear chain streaming applications on heterogeneous systems with failures
A survey of pipelined workflow scheduling: Models and algorithms
Reclaiming the energy of a schedule: Models and algorithms
Non-clairvoyant reduction algorithms for heterogeneous platforms
Non-linear divisible loads: There is no free lunch
Direct solvers for sparse linear systems
Push-relabel based algorithms for the maximum transversal problem
Constructing elimination trees for sparse unsymmetric matrices
Semi-matching algorithms for scheduling parallel tasks under resource constraints
Maximum cardinality bipartite matching algorithms on GPUs
Analysis of partitioning models and metrics in parallel sparse matrix-vector multiplication
On partitioning and reordering problems in a hierarchically parallel hybrid linear solver
UMPA: A Multi-objective, multi-level partitioner for communication minimization
A Partitioning-based divisive clustering technique for maximizing the modularity
Randomized matching heuristics with quality guarantees on shared memory parallel computers
On the minimum edge cover and vertex partition by quasi-cliques problems

Bilateral Contracts and Grants with Industry

Bilateral Contracts with Industry

Partnerships and Cooperations

National Initiatives
European Initiatives
International Initiatives
International Research Visitors

Dissemination

Scientific Animation
Teaching - Supervision - Juries

Bibliography

Publications of the year
References in notes

Home | Next next

2013 Team Activity Report

ROMA

Resource Optimization: Models, Algorithms, and scheduling

Research centre: Grenoble - Rhône-Alpes

In partnership with: CNRS, Université Claude Bernard (Lyon 1), Ecole normale supérieure de Lyon

In collaboration with: Laboratoire de l'Informatique du Parallélisme (LIP)

Field: Networks, Systems and Services, Distributed Computing

Theme: Distributed and High Performance Computing

Keywords: Scheduling, Parallel And Distributed Algorithms, Combinatorial Optimization, Exascale Systems, Fault Tolerance