ICS Cache Lab, Peking University.
This lab will help you understand the impact that cache memories can have on the performance of your C programs.
The lab consists of two parts. In the first part you will write a small C program (about 200-300 lines) that simulates the behavior of a hardware cache memory. In the second part, you will optimize a small matrix transpose function, with the goal of minimizing the number of cache misses.
For more information about this lab, please refer to cachelab.pdf.
It is recommended to do this lab on Ubuntu 22.04.
Before getting your hands dirty, a Linux program called valgrind will be doing you a great deal of things. You can type the following command to install valgrind.
sudo apt install valgrind
My score for this lab is as follows.
| Total score | Total misses | Misses (32×32) | Misses (64×64) | Misses (60×68) |
|---|---|---|---|---|
| 53.0 | 2682 | 256 | 1024 | 1402 |
The miss counts of transposing 32×32 and 64×64 matrices are theoretically optimal. The best miss count of transposing a 60×68 matrix in my class is 1382.
I have written a tool in parameter auto-search.cpp to automatically calculate the miss counts of different transpose methods. All you have to do is to write your transpose function in transpose_submit using load and store.
If you have some parameters in your transpose function, you can search for the best parameters automatically using this tool. I have given an example in parameter auto-search.cpp. Moreover, this tool enables more advanced approach, such as heuristic search and even neural networks.
Another powerful feature of this tool is that by setting print_codes = true, it prints out the corresponding C code of your method, which you can paste to trans.c and submit directly. This feature may come in handy if you are using more advanced approaches.