Multi-GPU and Multi-CPU Parallelization for Interactive Physics Simulations

Everton Hermann 1, 2 Bruno Raffin 2 François Faure 1 Thierry Gautier 2 Jérémie Allard 3
1 EVASION - Virtual environments for animation and image synthesis of natural objects
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
2 MOAIS - PrograMming and scheduling design fOr Applications in Interactive Simulation
Inria Grenoble - Rhône-Alpes, LIG - Laboratoire d'Informatique de Grenoble
3 ALCOVE - Collaborative interactive virtual environment
LIFL - Laboratoire d'Informatique Fondamentale de Lille, Inria Lille - Nord Europe, IRCICA
Abstract : Today, it is possible to associate multiple CPUs and multiple GPUs in a single shared memory architecture. Using these resources efficiently in a seamless way is a challenging issue. In this paper, we propose a parallelization scheme for dynamically balancing work load between multiple CPUs and GPUs. Most tasks have a CPU and GPU implementation, so they can be executed on any processing unit. We rely on a two level scheduling associating a traditional task graph partitioning and a work stealing guided by processor affinity and heterogeneity. These criteria are intended to limit inefficient task migrations between GPUs, the cost of memory transfers being high, and to favor mapping small tasks on CPUs and large ones on GPUs to take advantage of heterogeneity. This scheme has been implemented to support the SOFA physics simulation engine. Experiments show that we can reach speedups of 22 with 4 GPUs and 29 with 4 CPU cores and 4 GPUs. CPUs unload GPUs from small tasks making these GPUs more efficient, leading to a "cooperative speedup" greater than the sum of the speedups separatly obtained on 4 GPUs and 4 CPUs.
Complete list of metadatas


https://hal.inria.fr/inria-00502448
Contributor : François Faure <>
Submitted on : Thursday, July 15, 2010 - 1:07:30 AM
Last modification on : Thursday, February 21, 2019 - 10:52:48 AM
Long-term archiving on : Tuesday, October 23, 2012 - 10:20:44 AM

Files

europar10.pdf
Files produced by the author(s)

Identifiers

Citation

Everton Hermann, Bruno Raffin, François Faure, Thierry Gautier, Jérémie Allard. Multi-GPU and Multi-CPU Parallelization for Interactive Physics Simulations. Europar 2010 - 16th International Euro-Par Conference on Parallel Processing, Aug 2010, Ischia-Naples, Italy. pp.235-246, ⟨10.1007/978-3-642-15291-7_23⟩. ⟨inria-00502448⟩

Share

Metrics

Record views

1716

Files downloads

1317