Accelerating 'fields' by revamping the Cholesky Decomposition

The Geophysical Statistics project group within the Institute for Mathematics Applied to Geosciences (IMAGe) has been making use of Matrix Algebra on GPU and Multicore Architectures (MAGMA) to accelerate the Cholesky decomposition. The acceleration is motivated by a) Its frequent use in key computations in the spatial statistics R ‘fields’ package, b) Major bottleneck in ‘fields’ package execution and c) Operations involving big matrices make it suitable for parallelization. The Cholesky Decomposition was accelerated last summer using the MAGMA library. However, the performance of the accelerated version on multiple GPUs was observed to be unconventional - a) Execution time on multiple GPUs was higher in comparison to single GPU execution and b) Deep copy and in-place algorithms had opposite impacts on performance when executed on one and multiple GPUs. Our CPU and GPU profiling, conducted this summer, explains the unconventional behavior observed in the multi-GPU executions. The profiling provided insight to further accelerate the Cholesky Decomposition hierarchically– a) accelerating the underlying C function, b) reducing the function call overhead in R and c) optimizing the R environment. We were able to optimize the code and the environment to get a speedup greater than 75x (single precision) and 65x (double precision) for large matrices. We also found a potential way to improve the MAGMA functions by replacing the communications with direct device-to-device calls.

To Access Resource:

Questions? Email Resource Support Contact:

    UCAR/NCAR - Library


Resource Type publication
Temporal Range Begin N/A
Temporal Range End N/A
Temporal Resolution N/A
Bounding Box North Lat N/A
Bounding Box South Lat N/A
Bounding Box West Long N/A
Bounding Box East Long N/A
Spatial Representation N/A
Spatial Resolution N/A
Related Links N/A
Additional Information N/A
Resource Format PDF
Standardized Resource Format PDF
Asset Size N/A
Legal Constraints

Copyright Author(s). This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Access Constraints None
Software Implementation Language N/A

Resource Support Name N/A
Resource Support Email
Resource Support Organization UCAR/NCAR - Library
Distributor N/A
Metadata Contact Name N/A
Metadata Contact Email
Metadata Contact Organization UCAR/NCAR - Library

Author Ramakrishnaiah, Vinay
Kumar, Raghu
Paige, John
Hammerling, Dorit
Publisher UCAR/NCAR - Library
Publication Date 2015-08-01T00:00:00
Digital Object Identifier (DOI) Not Assigned
Alternate Identifier N/A
Resource Version N/A
Topic Category geoscientificInformation
Progress N/A
Metadata Date 2023-08-18T18:06:49.773181
Metadata Record Identifier edu.ucar.opensky::technotes:530
Metadata Language eng; USA
Suggested Citation Ramakrishnaiah, Vinay, Kumar, Raghu, Paige, John, Hammerling, Dorit. (2015). Accelerating 'fields' by revamping the Cholesky Decomposition. UCAR/NCAR - Library. Accessed 13 July 2024.

Harvest Source