Resource Allocation Constraints

Resource Allocation Constraints of Atlas 200T A2/Atlas 200I A2 Box16 Heterogeneous Subrack

For different service models, resource allocation constraints of training jobs on the Atlas 200T A2 Box16 heterogeneous subrack and Atlas 200I A2 Box16 heterogeneous subrack are as follows:

• The number of Ascend AI Processors allocated to a training job cannot be greater than the total number of Ascend AI Processors on a node.
• The number of Ascend AI Processors allocated to a training job can be 1 to 8, 10, 12, 14, or 16.
• If the number of Ascend AI Processors allocated to a training job is less than or equal to 8, select Ascend AI Processors in the same HCCS.
• If the number of Ascend AI Processors allocated to a training job is 10, 12, or 14, allocate the required Ascend AI Processors evenly to two rings. The relative physical addresses must be the same.
• If the number of Ascend AI Processors allocated to a training job is 16, all Ascend AI Processors of the node need to be allocated to this job.
• Resource allocation must comply with other constraints of the open source Volcano.