Rust-vs-OpenMP-Benchmarking-Code

This repository consists of different programs written for benchmarking Rust and C++.

The Generic Project Structure is as follows:

├───OpenMP
│   ├───<Benchmark Parameter 1>
│   │   ├───<benchmark_file_1.cpp>
│   │   └───<benchmark_file_2.cpp>
│   └───<Benchmark Parameter 2>
│       ├───<benchmark_file_1.cpp>
│       └───<benchmark_file_2.cpp>
├───Python - Test Case Builder
│   └───<python_helper_files.py>
├───Rust
│   ├───<Benchmark Parameter 1>
│   │   ├───<benchmark_file_1.rs>
│   │   └───<benchmark_file_2.rs>
│   └───<Benchmark Parameter 2>
│       ├───<benchmark_file_1.rs>
│       └───<benchmark_file_2.rs>
└───Test Inputs
    ├───<BenchmarkParameter_ProgramName_1.txt>
    └───<BenchmarkParameter_ProgramName_2.txt>

Steps to run the code on Crunchy1:

1. SSH into cims using: ssh <netID>@access.cims.nyu.edu
2. SSH into crunchy1 using: ssh crunchy1
3. Once the zip file is unzipped, and the contents are available, we will load the latest gcc module.
4. Load gcc-11.2 using: module load gcc-11.2
5. To compile matrix_multiplication.cpp, use: g++ -fopenmp -o <objectname> matrix_multiplication.cpp
6. To measure the time, use the following command: time ./<objectname> <square_matrix_length> <num_threads>
7. To compile merge_sort.cpp, use: g++ -fopenmp -o <objectname> merge_sort.cpp -lgomp
8. To measure the time, use the following command: time ./<objectname> <array_size> <num_threads>
9. To compile the Rust Code for matrix_multiplication, go into Rust\Scalability\matrix_multiplication.
10. Build release version using: cargo build --release
11. Run this release version using: cargo run --release <square_matrix_length> <num_threads>
12. To compile the Rust Code for merge_sort, go into Rust\Scalability\merge_sort.
13. Build release version using: cargo build --release
14. Run this release version using: cargo run --release <array_size> <num_threads>
15. To compile RuntimeOverhead.cpp, go into OpenMP\RuntimeOverhead, then use: g++ -fopenmp -o RuntimeOverhead RuntimeOverhead.cpp
16. To run RuntimeOverhead.cpp, use: ./RuntimeOverhead
17. To compile the Rust code for runtime overhead, go into Rust\RuntimeOverhead, then use: cargo build --release
18. To run this release version using: cargo run --release

Steps for referencing and running Programmability:

1. For this Project, we are comparing two languages: OpenMP and Rust.
2. For Programmability, all the code is present under the "Programmability" directory for both OpenMP and Rust:
   a. The "OpenMP/Programmability/" directory contains both the sequential and parallel codes for 
      Matrix Multiplication and Merge Sort are written in C++ and OpenMP respectively.
   b. The "Rust/Programmability/" directory contains both the sequential and parallel codes for
      Matrix Multiplication and Merge Sort are written in Rust and Rayon respectively.
   c. To run the files, we need to follow the same steps which are mentioned above.