dwymi02 / nheqminer

nheqminer (Zcash miner)

nheqminer

This version of nheqminer is a continuation of the development effort, with the hope of making this a better ZEC mining tool.

• This version of nheqminer supports both CPU and GPU (NVIDIA CUDA) ZEC mining

Table of contents

• Features
• Download
• Build
• Runtime Dependencies
• Usage
• Donations

Features

• Decent performance
• Windows support for now, provided by nicehash
• Linux support (tested on Ubuntu 18.04 LTS)
• Stratum protocol support

Download

• Binary releases: TBA
• Git tree: https://github.com/dwymi02/nheqminer.git
  • Clone with git clone https://github.com/dwymi02/nheqminer.git

Build

Dependencies

• Boost 1.62+

Windows

For the time being, Windows builds should be pulled from the original release, by NiceHash. Available here: https://github.com/nicehash/nheqminer/releases

Download and install:

• CUDA SDK (if not needed remove USE_CUDA_TROMP and USE_CUDA_DJEZO from nheqminer Preprocessor definitions under Properties > C/C++ > Preprocessor)
• Visual Studio 2013 Community: https://www.visualstudio.com/en-us/news/releasenotes/vs2013-community-vs
• Visual Studio Update 5 installed
• 64 bit version only

Open nheqminer.sln under nheqminer/nheqminer.sln and build. You will have to build ReleaseSSE2 cpu_tromp project first, then Release 7.5 cuda_tromp project, then select Release and build all.

Enabled solvers

• USE_CPU_TROMP
• USE_CPU_XENONCAT
• USE_CUDA_TROMP
• USE_CUDA_DJEZO

If you don't want to build with all solvers you can go to nheqminer Properties > C/C++ > Preprocessor > Preprocessor Definitions and remove the solver you don't need.

Linux

Working solvers CPU_TROMP, CPU_XENONCAT, CUDA_TROMP, CUDA_DJEZO

Dependencies

• CUDA -- 9.2
• Boost -- 1.62 (built using 1.65.1)
• build-essential -- 12.4
• CMake -- 3.10.2

General instructions

• NVIDIA CUDA, refer to installation instructions via URL: http://developer.download.nvidia.com/compute/cuda/9.2/Prod/docs/sidebar/CUDA_Installation_Guide_Linux.pdf

• The CUDA Toolkit can be downloaded from URL: https://developer.nvidia.com/cuda-downloads

• Get the latest nheqminer repository: git clone https://github.com/dwymi02/nheqminer.git

• Configure and build:

  • cd nheqminer/cpu_xenoncat/asm_linux
  • chmod +x fasm
  • sh assemble.sh
  • cd ../../../
  • mkdir build && cd build
  • cmake -DCUDA_CUDART_LIBRARY="/usr/local/cuda-9.2/lib64/libcudart.so" ../nheqminer
  • make -j $(nproc)

Dependencies

Runtime Dependencies required to start nheqminer

• NVIDIA CUDA toolkit.

  Please refer to the build section which provides the reference material for installing the NVIDIA CUDA toolkit

  For the time being the NVIDA CUDA toolkit is required even if you only wish to CPU mine ZEC coin. Development is working on providing a a CPU miner without the NVIDIA CUDA dependencies

Usage

Options

Parameters:
  -l [location] Stratum server:port
  -u [username] Username (wallet address)
  -a [port] Local API port (default: 0 = do not bind)
  -d [level]  Debug print level (0 = print all, 5 = fatal only, default: 2)
  -b [hashes] Run in benchmark mode (default: 200 iterations)
  -h    Print this help and quit

CPU settings
  -t [num_thrds]  Number of CPU threads
  -e [ext]  Force CPU ext (0 = SSE2, 1 = AVX, 2 = AVX2)

NVIDIA CUDA settings
  -ci   CUDA info
  -cd [devices] Enable CUDA mining on spec. devices
  -cb [blocks]  Number of blocks
  -ct [tpb] Number of threads per block

Example: -cd 0 2 -cb 12 16 -ct 64 128

When nheqminer is run without parameters, miner will utilize 75% of available logical CPU cores.

To see available parameters for nheqminer, from your command prompt enter the following command:

    nheqminer -h

Example to run benchmark on your CPU:

    nheqminer -b

Example to CPU mine, using your own wallet and worker id, on NiceHash USA server:

    nheqminer -l equihash.usa.nicehash.com:3357 -u YOUR_WALLET.YOUR_WORKER

Example to CPU mine, using your own wallet and worker id, on EU server, using 6 threads:

    nheqminer -l equihash.eu.nicehash.com:3357 -u YOUR_WALLET.YOUR_WORKER -t 6

Example to CPU mine, using your own wallet and worker id, on nanopool USA server:

    nheqminer -l zec-us-east1.nanopool.org:16666 -u YOUR_WALLET/YOUR_WORKER/[email protected] -p x -t [num_thrds]

Additional information about mining ZEC on nanopool please refer to https://zec.nanopool.org/help

Additional mining pools for your consideration available @ https://investoon.com/mining_pools/zec

Note: if you have a 4-core CPU with hyper threading enabled (total 8 threads) it is best to run with only 6 threads (experimental benchmarks shows that best results are achieved with 75% threads utilized)

Example to mine on your CPU as well on your CUDA GPUs with your own BTC address and worker1 on EU server, using 6 CPU threads and 2 CUDA GPUs:

    nheqminer -l equihash.eu.nicehash.com:3357 -u YOUR_WALLET.YOUR_WORKER -t 6 -cd 0 1

Donations

Donations are greatly appreciated and will be humbly accepted as reward for our labors ...

• BTC: 1HcQHQfhjmdTVBW6mYdsBP9esw3zns4S4u
• ETH: 0xef5aa02df61f419d5bf0d75bd5cf6ff3a3d83269
• ZEC: t1PrZg5dBhcm7MfC21dcrRMxpD9tfzJiqAD
• XMR: 46NX5s3K6ZX7cv4tF8C8moTUm3QpwNjjjarXfqUNsibxGc9aas6bSdmZCMDRgG9BWcPwoQrhJfGRCK6PAJTtBKT8Lpxct8F

nheqminer's Issues

Building nheqminer on an Ubuntu 21.10 VM

First of all, I should ask if building nheqminer on Ubuntu 21.10 is supported at all. Because I found that the CUDA toolkit is provided for 18.04 and 20.04 only. I installed the 20.04 CUDA into the 21.10 virtual machine and suspect that somebody is doing something nasty here. 😉

Nevertheless, I am going to report my issue just for the case it could reveal a bug.

This is my system information:
bunt-inxi.log

I slightly but minimally modified the build instructions.

• Cloned nheqminer in the ~/software/nheqminer directory.

• Created this script called nheqminer_build.sh in the ~/software directory:

  #!/bin/bash
  
  export LC_ALL=C
  export LANG=C
  export LANGUAGE=C
  cd nheqminer/cpu_xenoncat/asm_linux &&
  chmod +x fasm &&
  sh assemble.sh &&
  cd ../../../
  mkdir -p build && cd build &&
  rm -Rvf * &&
  cmake \
      -DCUDA_CUDART_LIBRARY="/usr/local/cuda-9.2/lib64/libcudart.so" \
      ../nheqminer &&
      # -DUSE_CUDA_DJEZO=0
  make clean &&
  make -j $(nproc)

• Then I ran the command ./nheqminer_build.sh 2>&1 | tee nheqminer_build.log the input of which follows:

  nheqminer_build.log

  The error reads as follows (excerpt from the above log):

  /home/pavel/software/nheqminer/cuda_djezo/../cpu_tromp/blake2/blake2.h:100:21: error: size of array element is not a multiple of its alignment
    100 |     blake2s_state S[8][1];
        |                     ^
  /home/pavel/software/nheqminer/cuda_djezo/../cpu_tromp/blake2/blake2.h:100:21: error: size of array element is not a multiple of its alignment
  /home/pavel/software/nheqminer/cuda_djezo/../cpu_tromp/blake2/blake2.h:108:21: error: size of array element is not a multiple of its alignment
    108 |     blake2b_state S[4][1];
        |                     ^
  /home/pavel/software/nheqminer/cuda_djezo/../cpu_tromp/blake2/blake2.h:108:21: error: size of array element is not a multiple of its alignment
  CMake Error at cuda_djezo_generated_equi_miner.cu.o.cmake:280 (message):
    Error generating file
    /home/pavel/software/build/cuda_djezo/CMakeFiles/cuda_djezo.dir//./cuda_djezo_generated_equi_miner.cu.o
  
  
  make[2]: *** [cuda_djezo/CMakeFiles/cuda_djezo.dir/build.make:84: cuda_djezo/CMakeFiles/cuda_djezo.dir/cuda_djezo_generated_equi_miner.cu.o] Error 1
  make[1]: *** [CMakeFiles/Makefile2:190: cuda_djezo/CMakeFiles/cuda_djezo.dir/all] Error 2
  make: *** [Makefile:149: all] Error 2
  

---

As you can see, I also played around with disabling the USE_CUDA_DJEZO flag (see the line commented out in the build script). Having that disabled, I ran basically into the same error in the blake2.h file.

read_until: End of file -- Reconnecting in 3 seconds

I have noticed that this problem occurs both on Windows and Linux. This problem was originally reported in [https://github.com/nicehash/nheqminer/issues/46], but no action was taken.

Following was taken from a running miner:
[00:38:20][0x00007f832f28fd00] Speed [15 sec]: 10.2 I/s, 19.5333 Sols/s
[00:38:56][0x00007f832cdc4700] stratum | Received new job #1534332102
[00:39:08][0x00007f832cdc4700] stratum | read_until: End of file
[00:39:08][0x00007f832cdc4700] stratum | Reconnecting in 3 seconds...

THANX(MKD).