zkwork_aleo_gpu_worker Download - zkwork_aleo_gpu

zkwork_aleo_gpu_worker

Other source code

cuda-0.2.4

Download

Aleo_prover

pool_address=aleo.hk.zk.work:10003
pool_address=aleo.asia1.zk.work:10003
pool_address=aleo.jp.zk.work:10003

Latest version

Nvidia: v0.2.4
AMD: v0.2.4

Release-notes

aleo_prover cuda-v0.2.4
- 6%-25% performance improvement on different epoch
aleo_prover ocl-v0.2.4
- The first release version for AMD GPUs
aleo_prover v0.2.3-fix
- Improve mining dashboard on HiveOS
- Add support for extra config on HiveOS like -g 0 -g 1 --custom_name zkwork_worker
aleo_prover v0.2.3
- ~10% performance improvement on different GPUs
- Improve network stability, miner will choose the best service among all provided pools, avoiding single-node issue
- Recommended to run with all available pools as follows

./aleo_prover --pool aleo.asia1.zk.work:10003 --pool aleo.hk.zk.work:10003 --pool aleo.jp.zk.work:10003 --address aleoxxx

aleo_prover v0.2.2
- 15%-25% performance improvement on different GPUs
- v0.2.2-lite size: 39M, v0.2.2-full size: 482M
- v0.2.2-lite takes about 2 mins to warmup during first run, and this process maybe longer on old CPUs
- v0.2.2-lite requires nvidia driver version >= 545.23.06
aleo_prover v0.2.1
- 75%+ performance improvement on low-end graphics cards like RTX2080Ti/50hx etc
- Run your own tests on high-end graphics cards like RTX4090 etc
aleo_prover v0.2.0
- Completely fix invalid share issue
- Reduce warmup time from about 1 min to less than 1 second, which improves mining efficiency by 3%~10%
- Fix issue that all GPUs stop mining when one GPU crashes
- Compatible with any CPUs
- Compatible with nvidia-driver version since 535.54.03 in our test, try for yourselves
aleo_prover v0.1.3-boost
- Network optimization version of v0.1.2-pre, use if you have network problems
aleo_prover v0.1.3-stable
- Network optimization version of v0.1.2-stable, use if you have network problems
aleo_prover v0.1.2-pre
- 20%+ performance improvement than v0.1.2-stable
- Longer warmup time than v0.1.2-stable
- Invalid share issue still under development with <1% chance of occurrence
- Recommended for CPU with more than 16 cores
aleo_prover v0.1.2-stable
- Completely solve CPU usage on old CPUs, compatible with any CPUs since this version
- Fix segment fault issue in some epochs
- Fix invalid share issue in some epochs
- Network reconnection improvemnt, remove unnecessary warmup after reconnection which brings higher mining performance
- Recommended for CPU with less than 16 cores
aleo_prover v0.1.1-hot/v0.1.0-hot
- Lower CPU usage
- Lower GPU memory usage
- Better network connectivity
- New service node: aleo.asia1.zk.work:10003
aleo_prover v0.1.1
- 70%+ performance improvement on RTX2080TI
- 20%+ performance improvement on 30 and 40 series
aleo_prover v0.1.0
- Performance improvement, hashrate doubled compare to v0.0.2
aleo_prover v0.0.2
- Slight performance improvement.
- Deprecate complicated pool connection options --tcp_server and --ssl_server.
- Introduce new connection flag --pool.
aleo_prover v0.0.1
- The first pre-mainnet version with 80w + hashrate on RTX 4090

GPU Performance

aleo_prover v0.0.1

8 * NVIDIA GeForce RTX 4090
epoch: 1825
epoch hash: ab15lkhs9yya5k7njqm75et7hkzarkrw09a970hv9506s28re4drsyqyspc7f

Name & Version	Performance
aleo_prover v0.0.1	650w

Requirements

OS Version: Ubuntu 20.04 +
Nvidia Driver Version: 535.54.03 +

Usage

The zk.work team <[email protected]>

Usage: aleo_prover [OPTIONS] --address <ADDRESS>

Options:
      --address <ADDRESS>          Specify this as a mining node, with the given aleo address
      --pool <SERVERS>             Specify the pool address that the worker is contributing to
      --verbosity <VERBOSITY>      Specify the verbosity of the node [options: 0, 1, 2, 3] [default: 0]
      --custom_name <CUSTOM_NAME>  Specify the custom name of this worker instance [default: sixworker]
  -g, --gpu_index <GPU_INDEXES>    Specify gpu index to solve puzzle, all gpus are used by default, eg. -g 0 -g 1 -g 2 ...
      --network <NETWORK>          Aleo Network(Testnet, Mainnet, Canary) , current default testnet [default: Testnet]
  -h, --help                       Print help
  -V, --version                    Print version

Mining Tutorial

Oneline command to start: ./aleo_prover --pool aleo.aisa1.zk.work:10003 --pool aleo.hk.zk.work:10003 --pool aleo.jp.zk.work:10003 --address aleoxxx --custom_name myprover

On Ubuntu

Get an Aleo wallet address on foxwallet, leowallet, puzzle wallet etc.
Download zkwork Nvidia miner: wget https://github.com/6block/zkwork_aleo_gpu_worker/releases/download/cuda-v0.2.4/aleo_prover-v0.2.4_cuda_full.tar.gz.
Download zkwork AMD miner: wget https://github.com/6block/zkwork_aleo_gpu_worker/releases/download/ocl-v0.2.4/aleo_prover-v0.2.4_ocl.tar.gz.
On Nvidia: tar -zvxf aleo_prover-v0.2.4_cuda_full.tar.gz && cd aleo_prover, on AMD: tar -zvxf aleo_prover-v0.2.4_ocl.tar.gz && cd aleo_prover.
Update your Aleo address in inner_prover.sh and set custom name for mining server.
Start mining with sudo chmod +x run_prover.sh && ./run_prover.sh.
Check mining log with tail -f prover.log.

On HiveOS

Choose your download url for your gpu.

DOWNLOAD_URL for Nvidia: https://github.com/6block/zkwork_aleo_gpu_worker/releases/download/cuda-v0.2.4/aleo_prover-v0.2.4_cuda_full.tar.gz
DOWNLOAD_URL for AMD: https://github.com/6block/zkwork_aleo_gpu_worker/releases/download/ocl-v0.2.4/aleo_prover-v0.2.4_ocl.tar.gz