Provided CPU baseline is located in baseline, while our GPU baseline is located in cuda_v1. To run the baselines,
-
Build the executable code:
make -
Run the cryptohash kernel:
sbatch run_test [32-character hash]
You can find the 32-bit character of your desired string by:
-
Build the executable code:
make wiki -
Run the wiki kernel:
./wiki [string]
The three optimizations mentioned in the report are located in cuda_v2, cuda_v3, and cuda_v4 respectively. cuda_v5 and cuda_v6 are two experimental optimizations that are not eventually adopted in the report. The process of running the optimized code is the same as running the baseline.
generate_testcase is used to generate the test sets, which are stored in Testcase. All folders ending in "_test" are used for testing. To run the testing codes,
-
Build the executable code:
make -
Run the cryptohash kernel:
sbatch run_test [number of passwords to be tested]