This script will perform matrix multiplication using CPU computation with NumPy and then compare it with GPU computation without CUDA using TensorFlow.
Let 
be a 
matrix and 
be a 
matrix. Then the 
matrix 
is the product of matrixes A and B and is denoted as 
, where the elements in the i th row and the j th column of the matrix 
can be expressed as
We can estimate the number of GigaFLOPS achieved for a specific matrix size and its corresponding execution time. GigaFLOPS is a common metric used to measure the performance of computations involving floating-point operations, such as matrix multiplication. Higher GigaFLOPS values generally indicate better performance, as it represents the number of operations the system can perform in one second.
num_flops = 2 * matrix_size3:**
This calculates the total number of floating-point operations (FLOPs) required to perform matrix multiplication for two square matrices of size matrix_size x matrix_size. The matrix multiplication algorithm typically involves matrix_size3 multiplications and matrix_size3 - 1 additions for each element in the resulting matrix. Since we are dealing with two matrices, the total number of FLOPs is doubled, hence 2 * matrix_size**3.
gflops = (num_flops / (execution_time * 1e9)):
This calculates the performance of the matrix multiplication operation in GigaFLOPS. execution_time represents the time taken to complete the matrix multiplication in seconds. The division num_flops / execution_time gives us the number of FLOPs per second, and dividing it by 1e9 converts it into GigaFLOPS.
