https://github.com/hunjixin/keccak_cu
fast keccak cuda implemetation
https://github.com/hunjixin/keccak_cu
Last synced: 13 days ago
JSON representation
fast keccak cuda implemetation
- Host: GitHub
- URL: https://github.com/hunjixin/keccak_cu
- Owner: hunjixin
- Created: 2024-06-18T09:04:52.000Z (11 months ago)
- Default Branch: master
- Last Pushed: 2024-09-14T13:49:38.000Z (8 months ago)
- Last Synced: 2024-09-14T23:46:32.668Z (8 months ago)
- Language: Go
- Homepage:
- Size: 1.75 MB
- Stars: 4
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome - hunjixin/keccak\_cu - fast keccak cuda implemetation (Go)
README
# keccak cuda hash
implemation of paper https://www.researchgate.net/publication/220849942_GPU_Implementation_of_the_Keccak_Hash_Function_Family.
also do some improvement with asm.
performance: test on 3060ti 43M hash/s
command
build ptx
```
cd gpulib
nvcc --ptx .\keccak.cu -o ./keccak.ptx
```
run bench
```
cd gputool
go build
./gputool --hash_per_thread 1000 --grid //default difficult nearly cannot find the nonce
./gputool --hash_per_thread 1000 --grid --difficulty 4443685057045916916839494277496125038705878205143482420972918669312 //expect to finish with 6 minitus for 1000m hash spped
```
run batch
```
cd gputool_batch
go build
./gputool_batch --hash_per_thread 1000 --grid --difficulty 4443685057045916916839494277496125038705878205143482420972918669312 //change to a completable difficulty
```