LEVEL 1
Optimization Removed - Shared Memory
Here the convolution texture code was used as a reference code which implements exactly the same algorithm as convolution seperable but uses texture memory instead of shared memory.
The removal of this optimization or I should say not using shared memory resulted in a drastic increase in the processing time.
The table below shows the processing time(ms) for 10 iterations of running the exe and getting convolved image.
The code can be downloaded from the download section.
