The previous article covered the basics of creating Flight Sheets for your workers, in this article we’ll dive deeper into fine-tuning your worker with Overclocking Profiles.
Disclaimer: Any settings that deviate from factory values may cause your system to become unstable. Use at your own risk.
In general terms, overclocking is pushing your hardware beyond it’s factory specs thus possibly improving its performance. An Overclocking Profile can also be used for other purposes, like downvolting — reducing your hardware’s power consumption. There are many applications for overclocking and we’ll try to cover some of them in this article.
Overclocking is a process of trial and error and requires patience to achieve the desired results. You should always let your hardware run with each set of new Overclocking Profiles for some time to see if those specific parameters will stay stable. Hardware can behave differently sometimes giving out unpredictable results. One way to see if your overclocking changes were successful is whether or not you start getting rejected shares.
Rejected shares
Rejected shares — shares that your miner worked but they were either not submitted in time to be included in the block or they had an error. The most common reason for a large amount of rejected shares is due to overclocking the GPU too much to the point where it produces faulty shares. The smaller the rate of rejected shares to accepted, the higher the efficiency of your GPU’s Overclocking Profile.
This said, most users leave their hardware at default settings as overclocking results may vary for each user. It also greatly depends on the actual components a manufacturer uses as some components are capable of being overclocked beyond spec and some simply don’t perform so great.
All Overclocking Profiles are created individually for each farm and can’t be shared across farms. You can either create a farm-wide profile or a worker specific profile. The worker specific settings will always override any farm-wide settings. For example, this way you can set the optimal desired settings for the whole farm and then fine-tune each rig individually.
To add a new profile, go to your farm’s Overclocking Profiles tab and click the Add OC Profile button. In the Save Overclocking as Template window give your profile a name and click Save.
Creating a new OC template
You’ve created a template that will later be used by specific workers or across your entire farm. This template can then be copied and modified for different miner or algorithm combinations.
The template we created will be used by all your GPUs, but the settings are different for Nvidia and AMD. You can have separate sets of settings for both GPU types within the same profile in case your rig is running both Nvidia and AMD GPUs and they will apply to each GPU type individually.
To edit your overclocking profile click the plus icon next to your the profile’s name and follow the instructions below depending on your GPU type.
Editing the OC template
Please note that you can use the most popular presets by going to the Popular Presets tab and finding the model of your GPU.
Before changing any settings you can run the nvidia-smi
command to find out your current GPU settings. To do this remotely from your dashboard, go to your worker and click the Run Command button in the toolbar at the top of the screen.
Toolbar
You can specify one value for all cards or list of values for each card separated by a space. For example:
150
— one value for all GPUs.
0 100 0 0 0 140
— means GPU0 = none, GPU1 = 100, etc.
Nvidia OC window
The Algo field is used to specify overclock settings for a specific algorithm. You can create a separate set of overclock setting for each algorithm individually or choose the Default Config that will apply to all algorithms, but can be overwritten by the algorithm settings themselves.
Here’s an example of an Overclocking Profile for an Nvidia GTX 1060 GPU.
Note, these values are for example purposes. Use at your own risk.
Nvidia OC example settings
Before changing any settings you can run the amd-info
command to find out your current GPU settings. To do this remotely from your dashboard, go to your worker and click the Run Command button in the toolbar at the top of the screen.
Toolbar
You can specify one value for all cards or list of values for each card separated by a space. For example:
150
— one value for all GPUs.
0 100 0 0 0 140
— means GPU0 = none, GPU1 = 100, etc.
AMD OC window
The Algo field is used to specify overclock settings for a specific algorithm. You can create a separate set of overclock setting for each algorithm individually or choose the Default Config that will apply to all algorithms, but can be overwritten by the algorithm settings themselves.
Here’s an example of an Overclocking Profile for an AMD RX 580 GPU.
Note, these values are for example purposes. Use at your own risk.
AMD OC example settings
Click Save and you’re done.
You can now apply the Overclocking Profile you’ve created by going to the Workers tab, then clicking the checkboxes on the left side of the workers to which you want to apply the Overclocking Profile. You should see a speedometer icon appear in the menu bar in the top right corner.
Applying the OC profile
Click it, and you should see a window with the number of workers you chose and a list of all your Flight Sheets.
List of OC profiles to apply to workers
Choose the Overclocking Profile you want to apply and click the Apply button. You should see a message telling you that the Overclocking command was sent to the worker. In a few seconds your workers should apply the changes.