This is the GIT repo for the work on building a parallel Tucker for combustion data.
For more information:
Grey Ballard
Alicia Klinvex
Hemanth Kolla
We would live to also acknowledge important contributions to this code by the following persons:
- Woody Austin - Author of the original code, built using an older tensor codebase from Sandia
- Tamara Kolda - Project lead on both original and second version of code
- Casey Battaglino - Important bug fix for computing TTMs in the case where the result is bigger than the input tensor
This code is still in development, but we welcome evaluation by friendly expert users. Please contact us if you have any questions, or submit an issue if you find a bug or wish to request a new feature.
MPI implementation (We use openMPI MPICH2, and MVAPICH2)
BLAS implementation
LAPACK implementation
C++11 or greater
Please see https://tensors.gitlab.io/TuckerMPI
G. Ballard, A. Klinvex, T. G. Kolda. TuckerMPI: A Parallel C++/MPI Software Package for Large-scale Data Compression via the Tucker Tensor Decomposition. ACM Transactions on Mathematical Software, Vol. 46, No. 2, Article ID 13, 31 pages, 2020. https://doi.org/10.1145/3378445
Parallel Tensor Compression for Large-Scale Scientific Data
Woody Austin, Grey Ballard, and Tamara G. Kolda
IPDPS'16, doi:10.1109/IPDPS.2016.67
The development of this software was supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, Applied Mathematics program and a Sandia Truman Postdoctoral Fellowship (LDRD funding). Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE–AC04–94AL85000.