Extreme-Scale Task-Based Cholesky Factorization Toward Climate and Weather Prediction Applications
Type
Conference PaperAuthors
Cao, QingleiPei, Yu
Akbudak, Kadir
Mikhalev, Aleksandr
Bosilca, George
Ltaief, Hatem

Keyes, David E.

Dongarra, Jack
KAUST Department
Extreme Computing Research CenterApplied Mathematics and Computational Science Program
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Office of the President
Date
2020-06-18Preprint Posting Date
2019Permanent link to this record
http://hdl.handle.net/10754/656453
Metadata
Show full item recordAbstract
Climate and weather can be predicted statistically via geospatial Maximum Likelihood Estimates (MLE), as an alternative to running large ensembles of forward models. The MLE-based iterative optimization procedure requires the solving of large-scale linear systems that performs a Cholesky factorization on a symmetric positive-definite covariance matrix—a demanding dense factorization in terms of memory footprint and computation. We propose a novel solution to this problem: at the mathematical level, we reduce the computational requirement by exploiting the data sparsity structure of the matrix off-diagonal tiles by means of low-rank approximations; and, at the programming-paradigm level, we integrate PaRSEC, a dynamic, task-based runtime to reach unparalleled levels of efficiency for solving extreme-scale linear algebra matrix operations. The resulting solution leverages fine-grained computations to facilitate asynchronous execution while providing a flexible data distribution to mitigate load imbalance. Performance results are reported using 3D synthetic datasets up to 42M geospatial locations on 130, 000 cores, which represent a cornerstone toward fast and accurate predictions of environmental applications.Citation
Cao, Q., Pei, Y., Akbudak, K., Mikhalev, A., Bosilca, G., Ltaief, H., … Dongarra, J. (2020). Extreme-Scale Task-Based Cholesky Factorization Toward Climate and Weather Prediction Applications. Proceedings of the Platform for Advanced Scientific Computing Conference. doi:10.1145/3394277.3401846Publisher
ACMConference/Event name
The Platform for Advanced Scientific Computing (PASC) ConferenceISBN
9781450379939Additional Links
https://dl.acm.org/doi/10.1145/3394277.3401846ae974a485f413a2113503eed53cd6c53
10.1145/3394277.3401846