Project-Runtime:Nvidia Professor Partnership

Team Runtime

Members

Overall Objectives

Designing Efficient Runtime Systems

Scientific Foundations

Runtime Systems Evolution

Application Domains

Panorama

Software

ForestGOMP
Hardware Locality
KNem
Marcel
NewMadeleine
PIOMan
MPICH2-NewMadeleine
Open-MX
PadicoTM
StarPU

New Results

Efficient scheduling of OpenMP threads on NUMA machines
High-performance memory migration in Linux
Communication Optimization over High Speed Networks
Multithreaded Communication Engine
High-performance message passing over generic Ethernet hardware
Scheduling over Heterogeneous Multicore Architectures
Optimization of the Sparse Direct Linear Solver PaStiX
High-Performance Intra-node Communications
Topology-aware High-Performance Computing

Contracts and Grants with Industry

PhD thesis co-supervised with CEA/DAM
Contract between INRIA and Myricom
Nvidia Professor Partnership

Other Grants and Activities

ANR projects
NEGST (NExt Grid Systems and Techniques)
COST Action IC0805 ComplexHPC (Open European Network for High-Performance Computing in Complex Environments)
PHC Pessoa MAE Grant
PEPPHER
Associate Team between Runtime and MPICH2 team
Collaboration with the Open MPI project
PHC Sakura MAE Grant
INRIA – EDF R&D
INRIA-UIUC joint laboratory for Petascale Computing
Expertise
Committees
Invitations

Dissemination

Reviews
Seminars
Teaching

Bibliography

Major publications
Publications of the year
References in notes

Inria / Raweb 2009
Presentation of the Project Runtime

Section: Contracts and Grants with Industry

Nvidia Professor Partnership

Participants : Cédric Augonnet, Raymond Namyst.

We established a Professor Partnership between Raymond Namyst and Nvidia in the context of our work on scheduling on heterogeneous multicore platforms (see Section 6.6 ). Nvidia provided us with multiple high-end GPUs so as to support our research.

Logo Inria