CacheTest

This archive is distributed in association with the journal Operations Research under the MIT License.

The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper This is a Template by T. Ralphs. The snapshot is based on this SHA in the development repository.

Important: This code is being developed on an on-going basis at https://github.com/tkralphs/ORJournalTemplate. Please go there if you would like to get a more recent version or would like support

Cite

To cite the contents of this repository, please cite both the paper and this repo, using their respective DOIs.

https://doi.org/10.1287/opre.2024.0000

https://doi.org/10.1287/opre.2024.0000.cd

Below is the BibTex for citing this snapshot of the respoitory.

@misc{CacheTest,
  author =        {T. Ralphs},
  publisher =     {Operations Research},
  title =         {{CacheTest}},
  year =          {2024},
  doi =           {10.1287/opre.2024.0000.cd},
  url =           {https://github.com/ORJournal/2024.0000},
}  

Description

The goal of this software is to demonstrate the effect of cache optimization.

Requirements

Hardware

The experiments were done on a single thread running on an Intel Core i7-1360P CPU with 32Gb memory (8 Gb should be sufficient, though).

Software

to build the software, a C++ compiler is needed. To build using the included makefile, a bash environment with make available is also required.

Building

In Linux, to build the version that multiplies all elements of a vector by a constant (used to obtain the results in Figure 1 in the paper), stepping K elements at a time, execute the following commands.

make mult

Alternatively, to build the version that sums the elements of a vector (used to obtain the results Figure 2 in the paper), stepping K elements at a time, do the following.

make clean
make sum

Be sure to make clean before building a different version of the code.

Results

Figure 1 in the paper shows the results of the multiplication test with different values of K using gcc 9.4 on an Ubuntu Linux box.

Figure 2 in the paper shows the results of the sum test with different values of K using gcc 9.4 on an Ubuntu Linux box.

Replicating

To replicate the results in Figure 1, do either

make mult-test

or

python scripts/test.py mult

To replicate the results in Figure 2, do either

make sum-test

or

python scripts/test.py sum

Ongoing Development

This code is being developed on an on-going basis at the author's Github site.

Support

For support in using this software, submit an issue.