Skip to content

Grand Chariot (Subsystem A): Resource group

Available resource groups on Grand Chariot (Subsystem A) are as follows.

Important

Available resource groups are different between the trial service (Dec. 2018) and the formal service (after Jan. 2019).

Trial service (Dec. 2018)

resource group limit of nodes limit of elapsed time limit of memory per node limit of simultaneous job submission per user limit of simultaneous job execution per user
ea 4 1 hour 338 GB 4 1
sa 8 6 hours 338 GB 4 2
ma 32 6 hours 338 GB 4 2
la 128 2 hours 338 GB 4 2
ha 256 2 hours 338 GB 4 2

Important

During the trial service, token is not consumed. Due to the system configuration, token is reduced but will be reset at the end of the trial service.

Trial and debug use (after Jan. 2019)

resource group limit of nodes limit of elapsed time limit of memory per node limit of simultaneous job submission per user limit of simultaneous job execution per user
ea 4 1 hour 338 GB 4 1

Important

No additional service applications are required for using the trail/debug resource group. (Token is not reduced when using this resource group.)

Shared node use (after Jan. 2019)

resource group limit of nodes limit of elapsed time limit of memory per node limit of simultaneous job submission per user limit of simultaneous job execution per user
sa 8 24 hours 338 GB 128 32
ma 32 24 hours 338 GB 128 32
la 128 72 hours 338 GB 128 32
ha 256 240 hours 338 GB 128 32

Important

Application of additional service is requierd for using the resource groups of the shared node use. Token (calculation time) is consumed depending on the product of (job elapsed time) and (the number of used nodes).

Occupied node use (after Jan. 2019)

resource group limit of nodes limit of elapsed time limit of memory per node limit of simultaneous job submission per user limit of simultaneous job execution per user
XXXXXXa (the number of nodes applied in additional service) 168 hours 338 GB unlimited unlimited

Here, "XXXXXX" is the group name.

limit of simultaneous job execution

The limit of simultaneous job execution is determined by the number of nodes applied in additional service; e.g. in the case that 8 nodes were applied in additional service, you can execute up to 8 jobs (if each job requires 1 node) at the same time.