Skip to content
No description, website, or topics provided.
C++ Cuda Shell
Branch: master
Clone or download
Latest commit f04b1ec Mar 19, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
AmoveoMinerGpuCuda.sln
AmoveoMinerGpuCuda.vcxproj
README.md
base64.cpp
base64.h
build.sh
main.cpp
poolApi.cpp
poolApi.h
sha256.cuh

README.md

AmoveoMinerGpuCuda

Amoveo Cryptocurrency Miner for Gpu work to be used with AmoveoPool2.com.

Tested Gpu Speeds:

  • Tesla V100: ??? Mh/s - Suggested BlockSize: 1024, Numblocks: 64
  • Tesla P100: ??? Mh/s - Suggested BlockSize: 192, Numblocks: 168
  • GTX1080 TI: 2900 Mh/s - Suggested BlockSize: 96, Numblocks: 168
  • GTX1060 6GB: 1300 Mh/s - Suggested BlockSize: 64
  • GTX1050: 550 Mh/s - Suggested BlockSize: 64
  • Tesla K80: 451 Mh/s - Suggested BlockSize: 128
  • 750TI: 323 Mh/s - Suggested BlockSize: 32

Default BlockSize is 64. Default NumBlocks is 96. Default SuffixMax is 65536.

  • Try various BlockSize setting values. Optimal setting for BlockSize is very personal to your system. Try BlockSize values like 96, 64, 32, or 128. A higher BlockSize is almost always better, but too high will crash the miner.
  • If you get too much OS lag, reduce the SuffixMax setting (at the cost of a some hash rate).
  • If your Memory Controller Load is constantly at 100%, you may want to try lowering your NumBlocks.

Best Settings from My Tests:

  • Gtx1060: BlockSize=64, NumBlocks=96
  • Gtx1050: BlockSize=64, NumBlocks=90
  • Tesla K80: BlockSize=128, NumBlocks=128
  • 750Ti: BlockSize=32, NumBlocks=64

Windows

Run Dependencies

  • Install Visual Studio 2015 (Community Edition is free.)
  • Install Cuda 9.1

Releases

Latest pre-built releases are here

Run

Example Usage:

AmoveoMinerGpuCuda.exe BPA3r0XDT1V8W4sB14YKyuu/PgC6ujjYooVVzq1q1s5b6CAKeu9oLfmxlplcPd+34kfZ1qx+Dwe3EeoPu0SpzcI=

Advanced Usage Template:

AmoveoMinerGpuCuda.exe <Base64AmoveoAddress> <CudaDeviceId> <BlockSize> <NumBlocks> <RandomSeed> <SuffixMax> <PoolUrl>
  • CudaDeviceId is optional an defaults to 0.
  • BlockSize is optional and defaults to 64.
  • NumBlocks is optional and defaults to 96.
  • RandomSeed is optional. Set this if you want multiple miners using the same address to avoid nonce collisions.
  • SuffixMax optional and defaults to 65536. Do NOT use anything higher than 65536. Lower numbers reduce OS lag and will reduce hash rate by a few percent.
  • PoolUrl is optional and defaults to http://amoveopool.com/work

Build

The Windows releases are built with Visual Studio 2015 with Cuda, RestCPP, boost, and openSSL.

Linux

Install dependencies

sudo apt-get install libcpprest-dev libncurses5-dev libssl-dev unixodbc-dev g++ git

Install CUDA9.1

wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-repo-ubuntu1604_9.1.85-1_amd64.deb
sudo apt-key adv —fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/7fa2af80.pub
sudo dpkg -i cuda-repo-ubuntu1604_9.1.85-1_amd64.deb
sudo apt update
sudo apt install cuda -y

Add these lines to the end of ~/.bashrc

export CUDA_HOME=/usr/local/cuda
export PATH="/usr/local/cuda/bin:$PATH"
export LD_LIBRARY_PATH="/usr/local/cuda/lib64:$LD_LIBRARY_PATH"

Build

git clone https://github.com/Mandelhoff/AmoveoMinerGpuCuda.git
cd ./AmoveoMinerGpuCuda
sh build.sh

Donations are welcome:

  • Amoveo: BPA3r0XDT1V8W4sB14YKyuu/PgC6ujjYooVVzq1q1s5b6CAKeu9oLfmxlplcPd+34kfZ1qx+Dwe3EeoPu0SpzcI=
  • Ethereum: 0x74e0aF0522024f2dd94F0fb9B82d13782ECCaaF5
You can’t perform that action at this time.