Lamellar is an asynchronous tasking runtime for HPC systems developed in RUST
A collection of benchmarks to test the functionality and performance of the Lamellar runtime (https://github.com/pnnl/lamellar-runtime)
Sept 2020: Update for Lamellar 0.2.1 release July 2020: Second alpha release Feb 2020: First alpha release
These benchmarks requires the following dependencies:
- Lamellar - now on crates.io At the time of release, Lamellar has been tested with the following external packages:
| GCC | CLANG | ROFI | OFI | IB VERBS | MPI | SLURM | LAMELLAR |
|---|---|---|---|---|---|---|---|
| 7.1.0 | 8.0.1 | 0.1.0 | 1.9.0 | 1.13 | mvapich2/2.3a | 17.02.7 | 0.2.1 |
The OFI_DIR environment variable must be specified with the location of the OFI installation. The ROFI_DIR environment variable must be specified with the location of the ROFI installation.
In the following, assume a root directory ${ROOT}
-
download Benchmarks to ${ROOT}/Benchmarks
cd ${ROOT} && git clone https://github.com/pnnl/lamellar-benchmarks -
download Lamellar to ${ROOT}/lamellar-runtime -- or update Cargo.toml to point to the proper location
cd ${ROOT} && git clone https://github.com/pnnl/lamellar-runtime -
see readmes in "histo", "triangle_count"
For use with distributed HPC systems, lamellar installation may require additional steps. See the Lamellar documentation for details.
The user may also set the number of worker threads via a environmental variable. See the Lamellare documentation for details.
- version 0.2:
- histo
- triangle count
- version 0.1:
- histo
- triangle count
Working on additional benchmarks
Ryan Friese - [email protected]
Roberto Gioiosa - [email protected]
Mark Raugas - [email protected]
This project is licensed under the BSD License - see the LICENSE.md file for details.
This work was supported by the High Performance Data Analytics (HPDA) Program at Pacific Northwest National Laboratory (PNNL), a multi-program DOE laboratory operated by Battelle.