Parallel Plotting Guide


What is parallel plotting?

In simplest terms, parallel plotting is the act of creating multiple plots simultaneously. Instead of creating your plots one at a time, you could theoretically create as many as you want!

While using the Graphical User Interface (GUI) you have a couple of options when it comes to parallel plotting, each has its own benefits and drawbacks. In this tutorial we will be going over how to set up a parallel plot using the queue method, and the parallel plotting feature built into the GUI.

What you need to know

There are a couple of things you need to be familiar with regarding your hardware before you can get started. First, you want to know how much ram your computer has. You need a minimum of 900MiBs per plot. Second, you need at least 1 core per plot from your CPU. An example, if you have 4 cores, you could plot 4 plots at 1 core each in parallel.

Note, As far as CPUs, going over 2 cores per plot seems at this moment to be diminished in returns, however that may vary from rig to rig. Same goes for ram, going over 6750MiB of ram does not seem to matter much in terms of speed.

The next thing on our list is ram. The default setting on the GUI is 3390 MiB. According to chia.net, the minimum amount of ram you should allocate is 900MiB.

In this tutorial, we will be working with 2 cores and 3390MiB of ram.

The last part we will focus on in this crash course to plotting will be your storage space. You need to be aware of how much space is available in your temporary (plotting) location, as well as your final (farming) location. You want to make sure you don’t go over the amount available when preparing your queues or parallels. For example, if a plot sized 101.4GiB requires 239GiB to complete, you should create 4 plots, maybe 5 depending on delay on a 1TB drive. Do note, once the plot has been completed, the plotting drive needs time to transfer it over to the farming drive before it removes the temporary files and frees up that space again.

Finally, we are getting started!

Parallel plotting is actually quite simple, however everyone has a different setup, so you will need to fine tune it until you get your optimal output!

Using the GUI, navigate to plots, then on the top right you will find a button labeled “add a plot” once you click on that it will open up a new section for creating your plots.

The Parallel Method

For this method, we will keep it simple. We are going to create two k=32 plots in parallel. Simply select the plot size from the dropdown in the first step.

Next, you select how many plots you want to create. In this case, we will be creating two. After you select the plots you want to create, select the radio button labeled “Plot In Parallel”. Once you do, you will notice a new option is revealed. This option allows you to set a delay and stagger your plots.

Why is this important?

You may want to stagger your plot because each step in the plotting process requires different components in order to complete. More information on that can be found here but the basic idea is that you want to try to start your next plot roughly around the same time your first plot finishes it’s phase, so that you have 4 plots in 4 different stages at any given time. Say you wanted to create 8 plots, stagger them so that when the first plot finishes, your fifth one starts.

Aside from that, you want to select your fast (usually M.2 NVMe SSD for best performance) as your temporary directory, and an HDD with plenty of storage as your permanent directory. After that, go ahead and click on Create Plot at the bottom and you will be on your way!

The Plot Queue Method

This method can be used either on it’s own, or in coordination with the Parallel method, but for now we will keep it simple.

The steps are largely the same, the only difference is, you will not get a delay option with this method. Once a plot in the queue is finished, it will then begin the next plot in the queue. For example, if we are doing 2 plots in this queue, when plot number one finishes, plot number two will begin. Now, you may be asking yourself, how is this a parallel plot? Well in short, its not, at least not yet.

Setting up a second queue

Once you have your first queue (default name is default) created, you can simply create a second queue. If you are following along, you should have 1 plot currently being plotted, with another one waiting and the GUI should have them labeled as “default” for their queue. So now, we are going to follow the exact same steps as the first queue, however in the Advanced Options dropdown, we will go ahead and rename the queue. Let’s go with “second”.

Once you start this plot process, you will notice that the first queue (default) is still running, but the second queue (second) has just begun. This essentially lets you plot in parallel by having queues named differently. The benefit of this process is that you can queue up more plots (as the GUI limits you on how many plots you can set up) so as opposed to having one plot queue of 29, you could have infinite plot queues set up. The only thing to watch out for is that you will have to start these queues at offset times manually.

So, which is better?

Well, that really depends! Every situation is different and you may have to play around with it to find out which method works best for you.

There you have it, now get to plotting and farm up some Chia!