Problem
Take the MIPS code for any DOT PRODUCT of two vectors.
Use MARS → Tools → Data Cache Simulator tool to measure the cache hit rate. Use Least Recently Used (LSU) replacement policy for all the below experiments
1. Direct-Mapped Cache: sweep the "Number of Blocks" and "Cache Block Size" values from 4 to 32 and obtain the Cache Hit Rate and Cache Size for each combination Create two 2D line graphs or 3D graphs to show the results of all measurements; one for Cache Hit Rate and another one for Cache Size
2. Fully Associative Cache: sweep the "Number of Blocks" and "Cache Block Size" values from 4 to 32 and obtain the Cache Hit Rate and Cache Size for each combination. Create two 2D line graphs or 3D graphs to show the results of all measurements; one for Cache Hit Rate and another one for Cache Size
3. N-way Set Associative Cache: sweep the "Number of Blocks" and "Cache Block Size" values from 4 to 32, and set size value from 1 to 4, and obtain the Cache Hit Rate and Cache Size for each combination (48 in total). Create graphs to show the results of all measurements in terms of Cache Hit Rate and Cache Size.
4. Choose the best cache placement policy and explain the reason for your choice.
5. Run the above experiments for the Random Block Replacement Policy and report the best case cache hit rate and its corresponding cache size.