LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS

Mohammad Alaul Haque Monil; Narasinga Rao Miniskar; Frank Y. Liu; Jeffrey S. Vetter; Pedro Valero-Lara

doi:10.1109/RSDHA56811.2022.00007

LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS

Mohammad Alaul Haque Monil, Narasinga Rao Miniskar, Frank Y. Liu, Jeffrey S. Vetter, Pedro Valero-Lara

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

7 Scopus citations

Abstract

In keeping with the trend of heterogeneity in high-performance computing, hardware manufacturers and vendors are developing new architectures and associated software stacks (e.g., libraries) to harness the best possible performance from commonly used kernels (e.g., linear algebra kernels). However, kernels tuned for one architecture are not portable to others. Moreover, the coexistence of different architectures in a single node makes orchestration difficult. To address these challenges, we introduce LaRIS, a portable framework for LAPACK functionalities. LaRIS ensures a separation between linear algebra algorithms and vendor-library kernels by using the IRIS run time and IRIS-BLAS library. Such abstraction at the algorithm level makes the implementation completely agnostic to the vendor library and architecture. LaRIS uses the IRIS run time to dynamically select the vendor-library kernel and suitable processor architecture at run time. Through LU factorization, we demonstrate that LaRIS can fully utilize different heterogeneous systems by launching and orchestrating different vendor-library kernels without any change in the source code.

Original language	English
Title of host publication	Proceedings of RSDHA 2022
Subtitle of host publication	Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	12-21
Number of pages	10
ISBN (Electronic)	9781665475686
DOIs	https://doi.org/10.1109/RSDHA56811.2022.00007
State	Published - 2022
Event	2022 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures, RSDHA 2022 - Dallas, United States Duration: Nov 13 2022 → Nov 18 2022

Publication series

Name	Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis

Conference

Conference	2022 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures, RSDHA 2022
Country/Territory	United States
City	Dallas
Period	11/13/22 → 11/18/22

Funding

This research used resources of the Oak Ridge Leadership Computing Facility at ORNL, which is supported by the US Department of Energy’s (DOE’s) Office of Science under Contract No. DE-AC05-00OR22725. This research was supported in part by the Exascale Computing Project (17-SC-20-SC), a collaborative effort of the DOE’s Office of Science and the National Nuclear Security Administration. This manuscript has been authored by UT-Battelle LLC under contract no. DE-AC05-00OR22725 with the US Department of Energy. The publisher, by accepting the article for publication, acknowledges that the US government retains a non-exclusive, paid up, irrevocable, world-wide license to publish or reproduce the published form of the manuscript, or allow others to do so, for US government purposes. The DOE will provide public access to these results in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

Funders	Funder number
U.S. Department of Energy
Office of Science	17-SC-20-SC, DE-AC05-00OR22725
National Nuclear Security Administration
UT-Battelle

Keywords

CPU
GPU
IRIS run time
LAPACK
LU factorization
Portability
Tasking
Tiling

Access to Document

10.1109/RSDHA56811.2022.00007

Cite this

Monil, M. A. H., Miniskar, N. R., Liu, F. Y., Vetter, J. S., & Valero-Lara, P. (2022). LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS. In Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis (pp. 12-21). (Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/RSDHA56811.2022.00007

Monil, Mohammad Alaul Haque ; Miniskar, Narasinga Rao ; Liu, Frank Y. et al. / LaRIS : Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS. Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis. Institute of Electrical and Electronics Engineers Inc., 2022. pp. 12-21 (Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis).

@inproceedings{2ac3ebf8beab45958a3111be05ad2855,

title = "LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS",

abstract = "In keeping with the trend of heterogeneity in high-performance computing, hardware manufacturers and vendors are developing new architectures and associated software stacks (e.g., libraries) to harness the best possible performance from commonly used kernels (e.g., linear algebra kernels). However, kernels tuned for one architecture are not portable to others. Moreover, the coexistence of different architectures in a single node makes orchestration difficult. To address these challenges, we introduce LaRIS, a portable framework for LAPACK functionalities. LaRIS ensures a separation between linear algebra algorithms and vendor-library kernels by using the IRIS run time and IRIS-BLAS library. Such abstraction at the algorithm level makes the implementation completely agnostic to the vendor library and architecture. LaRIS uses the IRIS run time to dynamically select the vendor-library kernel and suitable processor architecture at run time. Through LU factorization, we demonstrate that LaRIS can fully utilize different heterogeneous systems by launching and orchestrating different vendor-library kernels without any change in the source code.",

keywords = "CPU, GPU, IRIS run time, LAPACK, LU factorization, Portability, Tasking, Tiling",

author = "Monil, {Mohammad Alaul Haque} and Miniskar, {Narasinga Rao} and Liu, {Frank Y.} and Vetter, {Jeffrey S.} and Pedro Valero-Lara",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures, RSDHA 2022 ; Conference date: 13-11-2022 Through 18-11-2022",

year = "2022",

doi = "10.1109/RSDHA56811.2022.00007",

language = "English",

series = "Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "12--21",

booktitle = "Proceedings of RSDHA 2022",

}

Monil, MAH , Miniskar, NR, Liu, FY, Vetter, JS & Valero-Lara, P 2022, LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS. in Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis. Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis, Institute of Electrical and Electronics Engineers Inc., pp. 12-21, 2022 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures, RSDHA 2022, Dallas, United States, 11/13/22. https://doi.org/10.1109/RSDHA56811.2022.00007

LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS. / Monil, Mohammad Alaul Haque ; Miniskar, Narasinga Rao; Liu, Frank Y. et al.
Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis. Institute of Electrical and Electronics Engineers Inc., 2022. p. 12-21 (Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - LaRIS

T2 - 2022 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures, RSDHA 2022

AU - Monil, Mohammad Alaul Haque

AU - Miniskar, Narasinga Rao

AU - Liu, Frank Y.

AU - Vetter, Jeffrey S.

AU - Valero-Lara, Pedro

PY - 2022

Y1 - 2022

N2 - In keeping with the trend of heterogeneity in high-performance computing, hardware manufacturers and vendors are developing new architectures and associated software stacks (e.g., libraries) to harness the best possible performance from commonly used kernels (e.g., linear algebra kernels). However, kernels tuned for one architecture are not portable to others. Moreover, the coexistence of different architectures in a single node makes orchestration difficult. To address these challenges, we introduce LaRIS, a portable framework for LAPACK functionalities. LaRIS ensures a separation between linear algebra algorithms and vendor-library kernels by using the IRIS run time and IRIS-BLAS library. Such abstraction at the algorithm level makes the implementation completely agnostic to the vendor library and architecture. LaRIS uses the IRIS run time to dynamically select the vendor-library kernel and suitable processor architecture at run time. Through LU factorization, we demonstrate that LaRIS can fully utilize different heterogeneous systems by launching and orchestrating different vendor-library kernels without any change in the source code.

AB - In keeping with the trend of heterogeneity in high-performance computing, hardware manufacturers and vendors are developing new architectures and associated software stacks (e.g., libraries) to harness the best possible performance from commonly used kernels (e.g., linear algebra kernels). However, kernels tuned for one architecture are not portable to others. Moreover, the coexistence of different architectures in a single node makes orchestration difficult. To address these challenges, we introduce LaRIS, a portable framework for LAPACK functionalities. LaRIS ensures a separation between linear algebra algorithms and vendor-library kernels by using the IRIS run time and IRIS-BLAS library. Such abstraction at the algorithm level makes the implementation completely agnostic to the vendor library and architecture. LaRIS uses the IRIS run time to dynamically select the vendor-library kernel and suitable processor architecture at run time. Through LU factorization, we demonstrate that LaRIS can fully utilize different heterogeneous systems by launching and orchestrating different vendor-library kernels without any change in the source code.

KW - CPU

KW - GPU

KW - IRIS run time

KW - LAPACK

KW - LU factorization

KW - Portability

KW - Tasking

KW - Tiling

UR - http://www.scopus.com/inward/record.url?scp=85148247453&partnerID=8YFLogxK

U2 - 10.1109/RSDHA56811.2022.00007

DO - 10.1109/RSDHA56811.2022.00007

M3 - Conference contribution

AN - SCOPUS:85148247453

T3 - Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis

SP - 12

EP - 21

BT - Proceedings of RSDHA 2022

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 13 November 2022 through 18 November 2022

ER -

Monil MAH , Miniskar NR, Liu FY, Vetter JS , Valero-Lara P. LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS. In Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis. Institute of Electrical and Electronics Engineers Inc. 2022. p. 12-21. (Proceedings of RSDHA 2022: Redefining Scalability for Diversely Heterogeneous Architectures, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis). doi: 10.1109/RSDHA56811.2022.00007

LaRIS: Targeting Portability and Productivity for LAPACK Codes on Extreme Heterogeneous Systems by Using IRIS

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this