COL380 – Parallel and Distributed Programming

Assignments from the Parallel and Distributed Programming course (COL380) at IIT Delhi, taught by Prof. Subodh Kumar

This repository contains assignments that explore concepts in parallelism and distributed computing using C++, OpenMP, MPI, and CUDA.

---

📁 Assignments

Assignment 0: Analysis of Sequential Matrix Multiplication

• Explored the performance of different loop permutations in matrix multiplication using C++.
• Performance measured using perf and GNU Profiler (gprof).

---

Assignment 1: Sparse Matrix Multiplication using OpenMP

• Generation and multiplication of block sparse matrices using OpenMP in C++.
• Computed row statistics for analyzing sparse matrix behavior.

---

Assignment 2: Graph Centrality Computation using MPI

• Computed the most influential nodes per color in a colored graph using MPI.
• Implemented parallel processing of graph data and analyzed different MPI communication strategies.

---

Assignment 3: Matrix Rearrangement using CUDA

• Performed sorted rearrangement of matrices in parallel using CUDA on the GPU.
• Implemented in C++ with CUDA kernels for parallel acceleration.

---

Assignment 4: Sparse Matrix Multiplication on GPU with CUDA and MPI

• Optimized sparse matrix multiplication using a hybrid of CUDA, MPI, and OpenMP.
• Analyzed the scalability of the parallel implementation for varying input sizes.

---

Each assignment focuses on solving computational problems using different models of parallel and distributed computing. The implementations demonstrate practical knowledge in optimizing and analyzing performance on multicore CPUs and GPUs.