EWBF Cuda Equihash Miner

What is the Cuda Equihash miner from EWBF?

The Cuda miner by EWBF is considered one of the most popular miners for the Equihash algorithm (Zhash). This miner is compatible with Nvidia graphics processors and has two main competitors – miniZ and Funakoshi.

Currently, the latest version is EWBF Miner v0.6, which supports Equihash 144.5, 210.9, 192.7, and 96.5. There is also Cuda Zcash by EWBF, designed for mining coins such as Zcash and Zclassic. In this case, it is recommended to choose version 0.3.4b.

Cuda Equihash Miner v0.6 by EWBF has introduced many new features. The most important ones are listed below:

• Fixed an issue with the API socket not closing upon exit.
• Increased timeout during authorization and subscription.
• Algorithms 144.5, 192.7, and 96.5 have become faster.
• Algorithm 144.5 uses less VRAM.

Peculiarities

• The “solver” parameter is disabled.
• Solver for the equihash 200,9 algorithm has been removed (for the previous version of the equihash algorithm).
• Solvers for Equihash algorithms 192,7 and 144,5 have been added.
• Timestamps are now present on the screen and in the log file upon completion of the process.
• The argument –algo has been added, allowing the selection of an Equihash algorithm with two valid values: 144_5 and 192_7. This parameter is mandatory.
• Algorithms 192_7 and 144_5 require a higher video memory capacity. Meanwhile, power consumption and heat generation of graphics cards remain at a lower level compared to the classical Equihash algorithm.

Cuda Equihash EWBF miner performance

Below are the performance indicators of Nvidia graphics cards for the algorithm. By the way, algorithm 144.5 has become more popular thanks to Bitcoin Gold and BitcoinZ. It requires approximately 1.63 GB of video memory (VRAM) to operate.

• GeForce GTX 1050ti – approximately 18 S/s;
• GeForce GTX 1060 – approximately 30 S/s;
• GeForce GTX 1070 – approximately 45 S/s;
• GeForce GTX 1080 – approximately 50 S/s;
• GeForce GTX 1080ti – approximately 75 S/s.

Let’s move on to less popular models. The 210.9 is known for its association with Aion. It utilizes approximately 1.2 GB of VRAM.

• GeForce GTX 1050ti — approximately 75 S/s;
• GeForce GTX 1060 — approximately 130 S/s;
• GeForce GTX 1070 — approximately 190 S/s;
• GeForce GTX 1080 — approximately 210 S/s;
• GeForce GTX 1080ti — approximately 320 S/s.

The 192.7 is known for its association with ZERO. It utilizes approximately 2.5 GB of VRAM.

• GeForce GTX 1050ti – approximately 8 S/s;
• GeForce GTX 1060 – approximately 15 S/s;
• GeForce GTX 1070 – approximately 22 S/s;
• GeForce GTX 1080 – approximately 24 S/s;
• GeForce GTX 1080ti – approximately 35 S/s.

The most popular variant for 96.5 is MinexCoin (MNX). It utilizes approximately 1.6 GB of VRAM.

• GeForce GTX 1050ti – approximately 7.8 KS/s;
• GeForce GTX 1060 – approximately 14 KS/s;
• GeForce GTX 1070 – approximately 20 KS/s;
• GeForce GTX 1080 – approximately 25 KS/s;
• GeForce GTX 1080ti – approximately 37 KS/s.

How to setup Cuda Equihash EWBF miner

EWBF is very easy to set up. Here’s a basic configuration scheme for Equihash 144.5.

miner --algo 144_5 --server SERVER_ADDR --port PORT --user USER --pass PASSWORD

Here is an example of what a .bat file looks like for mining Bitcoin Gold on the 2Miners pool:

miner.exe --algo 144_5 --pers BgoldPoW --server https://btg.2miners.com --port 4040 --user YOUR_ADDRESS.RIG_ID --pass x

As an example, below, we provide a correct .bat file for Cuda Zcash EWBF. In this example, we are mining Zcash using the Equihash algorithm.

:restart
@echo turned off
TIMEOUT 5
miner --server zec.2miners.com --port 1010 --user YOUR_ADDRESS.RIG_ID --pass x --eexit 3 --pec --fee 0
ping 127.0.0.1 > zero
goto :restart

Advanced setting

Let’s look at the advanced configuration commands.

• --pec: Display the energy efficiency of graphics cards. The value 1 indicates to display, 0 indicates not to display.
• --cuda_devices: Use only specific graphics cards. Useful for farms with different types of graphics cards (e.g., AMD and Nvidia). Specify cards separated by a space (–cuda_devices 1 3 5). If nothing is specified, all GPUs will be used.
• --eexit: Program termination in case of any error. Option 1: close if unable to restart graphics cards, Option 2: close if connection to the server is lost, Option 3: close in both cases.
• --log: Create a program event log. A text log file will be created in the miner’s folder. Values: 0 – do not create, 1 – write only errors, 2 – record everything displayed on the console.
• --templimit: Specify the temperature at which a graphics card will stop participating in mining. After the temperature decreases, the mining process will resume.
• --intensity: Specify the program’s intensity. Values: from 1 to 64. For each graphics card, you can specify individual values separated by a space, for example: –intensity 64 60 59 64.

Problem solving

Swap file

So, if you encounter an “Out of memory error” message, increase the size of the page file in Windows (virtual memory). It is recommended to set the file size equal to the number of graphics cards multiplied by 6300 MB. For example, if you have 6 graphics cards, set the page file size to 6 * 6300 = 37800 megabytes.

To increase the page file size in Windows 10, follow these steps: open My Computer – go to Advanced system settings – select the Advanced tab – go to the Performance section – choose Settings – then go to Advanced – and finally, select Change. Set the required page file size according to your configuration. After making these changes, restart the system. Following these steps, the program should operate stably.

Hashrate is less than the old equihash

This is a feature of the new algorithm. For example, Nvidia 1080 TI produces 55-60 Sol/s instead of 750-790.

Versions and updates

v0.3.4b

• Improved speed by approximately 2%.
• Added an option for intensity, setting the maximum intensity.
• Added support for configuration files.
• Introduced the parameter -config to set the configuration file name.
• Added support for server recovery after failure through configuration files.
• Introduced new values for the API method getstat.
• Added an HTTP API accessible at http://127.0.0.1:42000/getstat for retrieving statistics.
• Included a simple web page with statistics; simply open the API address in a web browser (Experimental).
• Fixed several bugs.

v0.3.3b

• Added an option for the energy efficiency calculator (PEC).
• Added an option for developer fee control.
• For additional information, use the command miner --help.

v0.3.2b

• mproved speed by up to 2%, dependent on the graphics card model. Some cards may not show any speed changes. Reduced GPU memory usage. Enhanced stability.

v0.3.1b

• Increased speed by up to 4%, varying based on the graphics card model. Some cards may not exhibit any speed changes. Minor optimizations.

v0.3.0b

• Integrated NVML (NVIDIA Management Library) for temperature monitoring.
• Added the parameter –templimit to set the maximum GPU temperature.
• Introduced the –logfile option for specifying a custom log file name.
• Added the –api option for a simple JSON-based API.
• Option –solver now allows setting solvers for different devices.
• Optimized up to 2%, depending on the graphics cards.
• Refer to the Help.txt file for more detailed information.

v0.2.0b

• Added three additional solvers.
• Introduced a benchmark test that assesses each solver during miner startup and selects the fastest for each card.
• Added the option –solver for manually choosing a solver.
• Included the –eexit option to exit when the miner crashes, allowing for automation of restarts or other actions.
• Added the –logfile option for logging to a file.
• Implemented minor optimizations to solvers and the working cycle, resulting in a speed increase of 1–4%, depending on the card model.
• With the addition of new solvers, this version is expected to perform better with older graphics processors.
• For additional information on new parameters, refer to help -h for concise explanations and examples.

v0.1.0b

Increased speed by 2–5%. For older compute capability 2.x and 3.x cards, try version 0.0.6b as it may offer better performance.

v0.0.9b

• Increased speed by 10%.
• NOTE: This version may be unstable.
• For older compute capability 2.x and 3.x cards, consider trying version 0.0.6b, as it may provide better performance.

v0.0.8b

• “Cosmetic” solver optimizations. Increased speed by approximately 2%.
• Reduced the number of rejected shares.
• Implemented a new dynamic intensity system.