A set of benchmarks to stress GPGPU accelerated cluster of ARM SoCs (tested with Nvidia TX1)
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
CloverLeaf
HPL
Jacobi/src
TeaLeaf
caffe
README.md
paper.pdf

README.md

clusterSoCBench

clusterSoCBench is a collection of benchmarks ported to ARM that can be used to evaluate the CPU+GPGPU performance of an ARM-based cluster.

For more details please visit our publication available here as paper.pdf:

R. Azimi, T. Fox and S. Reda, "Understanding the Role of GPGPU-accelerated SoC-based ARM Clusters", to appear in IEEE Cluster 2017.

Build instructions

Copy the source code to all nodes and compile each benchmark using make. For HPL, we used OPENBLAS. For message passing library:, we used OpenMPI.

More information about the individual benchmarks can be found here:

TeaLeaf: [https://github.com/UK-MAC/TeaLeaf]

CloverLeaf: [https://github.com/UK-MAC/CloverLeaf]

Jacobi: [https://github.com/parallel-forall/code-samples/tree/master/posts/cuda-aware-mpi-example/src]

Caffe: [https://github.com/BVLC/caffe]

Contacts

If you use any of our ideas, please cite our paper. Our contact information is:
reza_azimi at brown dot edu, tyler_fox at brown dot edu or sherif_reda at brown dot edu