Replication Data for: Advancing the ChASE library through algorithm optimization and multi-level MPI parallelism (ICPSR doi:10.26165/JUELICH-DATA/OFQWGZ)

View:

Part 1: Document Description
Part 2: Study Description
Part 5: Other Study-Related Materials
Entire Codebook

(external link)

Document Description

Citation

Title:

Replication Data for: Advancing the ChASE library through algorithm optimization and multi-level MPI parallelism

Identification Number:

doi:10.26165/JUELICH-DATA/OFQWGZ

Distributor:

Jülich DATA

Date of Distribution:

2023-04-21

Version:

2

Bibliographic Citation:

Di Napoli, Edoardo, 2023, "Replication Data for: Advancing the ChASE library through algorithm optimization and multi-level MPI parallelism", https://doi.org/10.26165/JUELICH-DATA/OFQWGZ, Jülich DATA, V2

Study Description

Citation

Title:

Replication Data for: Advancing the ChASE library through algorithm optimization and multi-level MPI parallelism

Subtitle:

SC23 Artifact Description (AD) dataset

Identification Number:

doi:10.26165/JUELICH-DATA/OFQWGZ

Authoring Entity:

Di Napoli, Edoardo (Forschungszentrum Jülich)

Other identifications and acknowledgements:

Di Napoli, Edoardo

Producer:

Di Napoli, Edoardo

Schleife, Andre

Software used in Production:

FLEUR

Software used in Production:

UIUC BSE code

Grant Number:

slai

Distributor:

Jülich DATA

Access Authority:

Di Napoli, Edoardo

Access Authority:

Wu, Xinzhe

Depositor:

Di Napoli, Edoardo

Date of Deposit:

2023-04-18

Date of Distribution:

2023-04-21

Study Scope

Keywords:

Mathematical Sciences, Physics, Eigenproblem, Condensed Matter Physics, Numerical Linear Algebra

Abstract:

This dataset includes a set of matrices generated by Condensed Matter simulation software which were used to test the ChASE library in the manuscript bearing the same title as the dataset, and submitted to SC23 conference. Each Hermitian matrix is stored in double precision as a separate binary file containing the whole matrix entries in column major order. This is the list of matrices and their sizes: NaCl 9k: n = 9273 -- AuAg 13k: n = 13379 -- TiO2 12k: n = 12455 -- TiO2 29k: n = 29528 -- In2O3 76k: n = 76887 -- In2O3 115k: n = 115459 -- HfO2 62k: n = 62681 -- HfO2 76k: n = 76674

Kind of Data:

other

Notes:

The data is accessible at https://datapub.fz-juelich.de/sdlqm/Materials_Matrices_SC23/

Methodology and Processing

Sources Statement

Data Access

Notes:

CC0 Waiver

Other Study Description Materials

Related Publications

Citation

Identification Number:

10.1145/3624062.3624249

Bibliographic Citation:

Xinzhe Wu, Edoardo Di Napoli. "Advancing the distributed Multi-GPU ChASE library through algorithm optimization and NCCL library." Proc. of the SC '23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis. Pages: 1688 - 1696

Other Study-Related Materials

Label:

SHA256SUMS.txt

Text:

Checksums for each matrix in the dataset

Notes:

text/plain