This repository features implementation of CC_load algorithm which tries to solve the following problem using heuristics in polynomial time .

*Given a task graph of modules, the problem of finding an optimal clustering of modules is an NP-Complete problem *

The repository also features the implementation of optimal algorithm which tries to compute execution time for all the allocations possible for the given graph.

Requirements - C++ 17 , GCC / Clang compiler , Cmake

Running the code -

In the CCLoad-implementation folder run the following command

mkdir build
cd build
cmake ..
make
./exe

The algorithm takes a task graph as input and tries to find the T_OPT (optimal time) and T_SUBOPT (sub-optimal time) for the given graph .

Graph is described as a adjaceny matrix G and a weight array W

let the nodes of graph be N

• W[i] means the computation time for ith node is W[i]
• G[i][j] means the communication time from ith node the jth , if G[i][j] > 0