Network Monitor
What is the Network Monitor?
The network monitor is a feature that allows Chefs to have a birds eye view on the state of demand for various Graphics Cards (GPUs) and their performance on the network, giving insights into what you can expect as a Chef owning such GPUs.
What information is held on the Network Monitor?
From left to right on the table, here is each entry and what it represents:
-
GPU:
- This part of the table lets you know for each row of the table which GPU is being concerned by the data, as well as the hourly rate range that this GPU can expect while Chopping a job from low to high.
-
Recommended Specs:
- This item lets you know what specs are currently most seeked out to be paired with your GPU in order to maximize your earnings in terms of RAM and storage. Worry not, if you don't meet this criteria it does not mean that you will not get jobs, though these recommended specs are a representation of what our clients regularly request for your GPU.
-
Demand:
- Demand for the current GPU, with values being Low, Moderate, or High. A GPU with Low demand means that there are many more Chefs with the requested hardware than there are jobs available. If demand is High, then we expect that a Chef with this type of hardware should get jobs fairly consistently.
-
Average Earning Rate, hourly:
- Tracks the average earning rate for all Chefs that have this GPU per hour.
-
Average Earning Rate Top 25% hourly:
- Tracks the cumulative average earnings per hour for the top 25% of earners that have this GPU. This column represents what our top earners for this GPU are earning per hour.
Why not show the total number of machines in the network?
The Salad network is made up of machines that may be available during several periods throughout the day. So if a job requires a GPU for 24 hours it may take several machines with that GPU to make up those 24 hours. This means that a count of machines in the network would not be enough to answer questions like "is it worth it to bring my hardware to Salad right now?".
Instead we're showing you a compound demand signal that takes into account total utilization of that GPU in the network along with average earnings and average running time for the last 24 hours.
What do the demand levels mean:
The demand for any GPU in the table (each line) can be marked with one of the three following values:
-
Low
- A GPU marked with Low demand suggests that there are currently more Chefs that are actively Chopping and waiting for a job than there are jobs available.
-
Moderate
- A GPU marked with Moderate demand suggests that jobs are available in the network for this GPU class, but there may still be some periods where your machine remains idle since there are more Chefs Chopping than there are jobs.
-
High
- A GPU marked with High demand suggests that Chefs with this type of hardware should see minimal gaps in their earning experience, and the machines should be able to get jobs fairly consistently.
Note that even though a GPU type may be in Low or Moderate supply, there are still jobs available for these machines. Chefs can expect there to be some gaps though in the earning experience after a job ends and the machine looks for the next job it can run on the network.
Why does higher demand not always equal higher payouts?
Salad offers GPUs at different priority tiers, each commanding a different hourly rate, so job availability (demand) is not tied to what those jobs are paying. Chefs should make an informed decision using all available data.
How can I get into those top 25% of earners?
Increasing your chopping time is the easiest way to improve your earning potential, especially if you achieve Star Chef status. Beyond that, follow any requirements or suggestions in your Chopping Power checklist inside the app.
Why do some GPUs have average earning rates higher than the reference hourly rates?
The reference hourly rates provided for each GPU factor in just the earning rate for the GPU, without extras for RAM or CPU. While these other hardware resources rarely contribute significantly to the earning rate of a machine, certain GPUs at certain times may have enough jobs requiring significantly higher resources which can cause the average earning rates to go above the reference hourly rates.
Last updated on