Compute Units #
Compute units allow specifying the amount of machine resources available for a single participant. Some tests require more computing power to run smoothly, but some need less than usual. Picking the correct compute unit value is critical for a smooth test run and cost optimisation.
Each full compute unit corresponds to 0.5 CPU core and 1 GB of system memory. In total Loadero currently offers 5 compute unit settings.
|Compute unit||CPU cores||Memory|
When using video or audio feed, we would suggest taking at least G2 compute unit participants that will provide 1 CPU core and 2 GB of system memory. For more resource demanding actions like group calls or peer-to-peer calls most of the time G4 or higher compute unit is necessary.
Default compute unit value for any participant is G1 (0.5 CPU core and 1 GB System memory) unless otherwise specified by the user.
Remember that group count and participant count will multiply used compute units. For example, when creating a participant with G2 compute unit and participant count is set to 10, it will result in total 20 (2 * 10) compute units used.
Subscription limits #
Each project subscription plan has some sort of limitation on how many compute units can be used in a single test and how many in total in the billing period.
All paid subscriptions (except Enterprise) can run max 200 000 compute units in a month. In case a higher compute unit limit is needed, contact support to apply for an Enterprise subscription.
|Subscription plan||Max compute units in a test||Base included compute units (per month)||Highest allowed compute unit size|
|Pay as you go: Essential||20 000||-||G2|
|Pay as you go: Ultimate||50 000||-||G6|
|Monthly||10 000||10 000||G6|
|Yearly||10 000||10 000||G6|
Check out our pricing page for more detailed subscription plan comparison.
For example, a test running 100 participants with compute unit value set to G4, will result in 400 compute unit usage. Using the full limit of the monthly subscription plan (10 000 compute units), it’s possible to launch from 1 666 G6 participants (1 666 * 6 = 9996) to 20 000 G0.5 participants (20 000 * 0.5 = 10 000) in a single run, depending on the compute unit configuration.