Shot level parallelization of a seismic inversion code using PVM
- Rice Univ., Houston, TX (United States)
- Inria, Le Chesnay (France)
This paper presents experience with parallelization using PVM of DSO, a seismic inversion code developed in The Rice Inversion Project. It focuses on one aspect: trying to run efficiently on a cluster of 4 workstations. The authors use a coarse grain parallelism in which they dynamically distribute the shots over the available machines in the cluster. The modeling and migration of their code is parallelized very effectively by this strategy; they have reached a overall performance of 104 Mflops using a configuration of one manager with 3 workers, a speedup of 2.4 versus the serial version, which according to Amdahl`s law is optimal given the current design of their code. Further speedup is currently limited by the non parallelized part of their code optimization, linear algebra and i(o).
- OSTI ID:
- 80193
- Report Number(s):
- CONF-941015--
- Country of Publication:
- United States
- Language:
- English
Similar Records
Parallel computation of genetic likelihoods using CRI-MAP, PVM, and a network of distributed workstations
Model of parallel performance. Final report, July-December 1988