This repo contains reference CPU and CUDA miners for the AION testnet. The reference miners have been specifically designed to be used in conjunction with the provided reference mining pool, future versions will be adapted to join public pools.

Note The precompiled versions of the CPU and GPU miners have been compiled with AVX CPU features enabled. AVX features may not be supported on all CPU models, this section outlines details to enable and disable AVX features during the build process. In order to use non-AVX versions of the miners, the miners must be built from source.

• Ubuntu 16.04+
• CUDA 9.0 compatible driver (NVIDIA 387.34+) for GPU based mining.

• Download a pre-built miner binary from the aion_miner (release) page.
• Unpack the miner.

  tar xvjf aionminer-<TYPE>-<VERSION>.bz2
  

• Note: When using the pre-built miner the CPU executable will be aionminer-cpu while the CUDA executable will be aionminer. When building from source the executable will be called aionminer for both CPU and CUDA builds.

• python-dev 2.7.11+
• Boost 1.65.1 (Higher versions are not currently supported by CMake at this time). The following libraries must be built:
  • system
  • log
  • date_time
  • filesystem
  • thread
  • program_options
• cmake 3.10.0

Windows builds are not currently supported.

Work in progress. Working solvers CPU_TROMP, CUDA_TROMP

• Install Boost 1.65.1

  wget https://dl.bintray.com/boostorg/release/1.65.1/source/boost_1_65_1.tar.bz2
  tar xf boost_1_65_1.tar.bz2
  cd boost_1_65_1
  ./bootstrap.sh --with-libraries=system,log,date_time,filesystem,thread,program_options
  sudo ./b2 install

• Install CUDA SDK v9 (make sure you have cuda libraries in LD_LIBRARY_PATH and cuda toolkit bins in PATH)

  • example on Ubuntu:

  LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/local/cuda-9.0/lib64:/usr/local/cuda-9.0/lib64/stubs"
  PATH="$PATH:/usr/local/cuda-9.0/"
  PATH="$PATH:/usr/local/cuda-9.0/bin"

• Use Boost 1.65.1+ (if it is not available from the repos you will have to download and build it yourself)

• CMake v3.10 (if it is not available from the repos you will have to download and build it yourself)

• Currently support only static building (CUDA_TROMP are enabled by default, check CMakeLists.txt in aion_miner root folder)

• Before proceeding ensure the correct solver is enabled

  • By default the CUDA-Tromp is enabled however if CUDA is not avalible this must be switch to CPU-Tromp before proceeding. The Switch Solvers outlines how to switch between solver implementations.

• To compile the miner; open the terminal and run the following commands:

  git clone https://github.com/aionnetwork/aion_miner.git
  cd aion_miner
  mkdir build && cd build
  cmake ../aion_reference_miner
  make

• Navigate to the root of the project directory
• Open the CMakelists.txt file
• Toggle ON/OFF settings on lines 7 and 8 to enable/disable solvers. Rebuild project after each toggle.
• Only a single solver should be used at a time.

• Navigate to the root of the project directory
• Open the CMakelists.txt file
• Navigate to the Linux section at line 21.
• To disable AVX features and enable SSE2:
  • Comment lines 26 and 27 by places a # in from of the lines.
  • Uncomment lines 23 and 24 by removing # at the start of the lines.
• To enable AVX features and disable SSE2:
  • Uncomment lines 26 and 27 by places a # in from of the lines.
  • Comment lines 23 and 24 by places a # in from of the lines.

Parameters:

  -h [ --help ]                   Print help messages
  -l [ --location ] arg           Stratum server:port
  -u [ --username ] arg           Username (Aion Addess)
  -a [ --apiPort ] arg            Local port (default 0 = do not bind)
  -d [ --level ] arg              Debug print level (0 = print all, 5 = fatal 
                                  only, default: 2)
  -b [ --benchmark ] [=arg(=200)] Run in benchmark mode (default: 200 
                                  iterations)
CPU Parameters:
  -t [ --threads ] arg            Number of CPU threads
  -e [ --ext ] arg                Force CPU ext (0 = SSE2, 1 = AVX, 2 = AVX2)
CUDA Parameters:
  --ci                            Show CUDA info
  --cv arg                        CUDA solver (0 = djeZo, 1 = tromp, default=1)
  --cd arg                        Enable mining on spec. devices
  --cb arg                        Number of blocks (per device)
  --ct arg                        Number of threads per block (per device)

Note All CUDA parameters may also be used with a single "-".

Run AION CPU miner with 4 threads connecting to a mining pool running locally, listening on port 3333 for incoming connections.

./aionminer -t 4 -l 127.0.0.1:3333

Example to run benchmark on your CPU (Single thread):

./aionminer -b -t 1

Example to run benchmark on your CPU using 300 iterations (Single thread):

./aionminer -b 300 -t 1

Run AION CUDA miner with 64 blocks, 64 threads per block on device 0 using solver version 1 (CUDA Tromp)

./aionminer -cd 0 -cv 1 -cb 64 -ct 64

Example to run benchmark on your GPU (GPU 0, CUDA-Tromp solver):

./aionminer -cd 0 -cv 1 -cb 64 -ct 64 -b