Skip to content

ecrc/hicma-x

Repository files navigation

HiCMA-PaRSEC

Table of Contents

Overview

HiCMA-PaRSEC, i.e., Hierarchical Computations on Manycore Architectures (HiCMA) powered by Parallel Runtime Scheduler and Execution Controller (PaRSEC), combines the linear algebraic contributions of mixed-precision and low-rank computations. The framework features adaptive memory management and precision selection capabilities to optimize performance and accuracy for extreme-scale scientific computing applications.

HiCMA-PaRSEC poster

Features

Core Computational Capabilities

  • Matrix Operations: Matrix multiplication, Cholesky factorization, rank-k updates, etc.
  • Tile Low-Rank (TLR) Computations: Hierarchical matrix representations for sparse problems
  • Mixed Precision Support: Double, single, half, and FP8 precision arithmetic
  • Adaptive Memory Management: Dynamic memory allocation and precision selection
  • GPU Acceleration: CUDA and HIP support for heterogeneous computing

Runtime and Parallelism

  • PaRSEC Dynamic Runtime: Task-based parallel execution with automatic load balancing
  • Multi-Platform Support: Shared and distributed-memory environments
  • Multi-GPU Systems: Scalable GPU acceleration across multiple devices
  • Extreme Scale: Optimized for exascale computing applications

Application Domains

  • Climate Emulator: Exascale climate emulator with spherical harmonic transforms
  • Geospatial Modeling: Spatial statistics and environmental applications
  • Genomics: Genetic analysis
  • Hamming Distance: Hamming distance computations
  • Scientific Computing: 3D unstructured mesh deformation (e.g., SARS-CoV-2 modeling)

Building

For detailed instructions, please refer to BUILD.md.

Testing

See TESTS.md for detailed examples and parameter configurations.

Literature

2025

  • R. Alomairy, S. Abdulah, Q. Cao, D. Keyes, and H. Ltaief. "Sustainably Modeling a Sustainable Future Climate." IEEE High Performance Extreme Computing Conference, 2025.

  • R. Alomairy, Q. Cao, H. Ltaief, and D. Keyes. "Scalable Low-Rank Solvers for Large-Scale 3D Mesh Deformation Using Global and Compact Support RBF Kernels." IEEE International Symposium on Cluster, Cloud, and Internet Computing (CCGRID, TCSC SCALE Challenge Award, Finalist), 2025.

  • R. Alomairy, Q. Cao, H. Ltaief, D. E. Keyes, and A. Edelman. "Scalable Hamming Distance Computation Using Accelerated Matrix Transformations." ISC High Performance, 2025.

2024

  • S. Abdulah, A. H. Baker, G. Bosilca, Q. Cao, S. Castruccio, M. G. Genton, D. E. Keyes, Z. Khalid, H. Ltaief, Y. Song, G. L. Stenchikov, and Y. Sun. "Boosting Earth System Model Outputs And Saving PetaBytes in their Storage Using Exascale Climate Emulators." International Conference for High Performance Computing, Networking, Storage and Analysis (SC), ACM Gordon Bell Prize for Climate Modelling, 2024.

  • H. Ltaief, R. Alomairy, Q. Cao, J. Ren, L. Slim, T. Kurth, B. Dorschner, S. Bougouffa, R. Abdelkhalek, and D. E. Keyes. "Toward Capturing Genetic Epistasis From Multivariate Genome-Wide Association Studies Using Mixed-Precision Kernel Ridge Regression." International Conference for High Performance Computing, Networking, Storage and Analysis (SC), ACM Gordon Bell Prize Finalist, 2024.

2023

  • Q. Cao, S. Abdulah, H. Ltaief, M. G. Genton, D. E. Keyes, and G. Bosilca. "Reducing Data Motion and Energy Consumption of Geospatial Modeling Applications Using Automated Precision Conversion." IEEE International Conference on Cluster Computing (CLUSTER), 2023.

2022

  • Q. Cao, S. Abdulah, R. Alomairy, Y. Pei, P. Nag, G. Bosilca, J. Dongarra, M. G. Genton, D. E. Keyes, H. Ltaief, and Y. Sun. "Reshaping geostatistical modeling and prediction for extreme-scale environmental applications." International Conference for High Performance Computing, Networking, Storage and Analysis (SC), ACM Gordon Bell Prize Finalist, 2022.

  • S. Abdulah, Q. Cao, Y. Pei, G. Bosilca, J. Dongarra, M. G. Genton, D. E. Keyes, H. Ltaief, and Y. Sun. "Accelerating geostatistical modeling and prediction with mixed-precision computations: A high-productivity approach with parsec." IEEE Transactions on Parallel and Distributed Systems (TPDS), 2022.

  • Q. Cao, R. Alomairy, Y. Pei, G. Bosilca, H. Ltaief, D. Keyes, and J. Dongarra. "A framework to exploit data sparsity in tile low-rank Cholesky factorization." IEEE International Parallel & Distributed Processing Symposium (IPDPS), 2022.

2021

  • Q. Cao, Y. Pei, K. Akbudak, G. Bosilca, H. Ltaief, D. Keyes, and J. Dongarra. "Leveraging parsec runtime support to tackle challenging 3d data-sparse matrix problems." IEEE International Parallel and Distributed Processing Symposium (IPDPS), 2021.

2020

  • Q. Cao, Y. Pei, K. Akbudak, A. Mikhalev, G. Bosilca, H. Ltaief, D. Keyes, and J. Dongarra. "Extreme-scale task-based Cholesky factorization toward climate and weather prediction applications." ACM Platform for Advanced Scientific Computing Conference (PASC), 2020.

2019

  • Q. Cao, Y. Pei, T. Herault, K. Akbudak, A. Mikhalev, G. Bosilca, H. Ltaief, D. Keyes, and J. Dongarra. "Performance analysis of tile low-rank Cholesky factorization using parsec instrumentation tools." IEEE/ACM International Workshop on Programming and Performance Visualization Tools (ProTools at SC), 2019.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

No packages published