Supports integrated job scheduling for both Kubernetes native workloads and mainstream computing frameworks (such as TensorFlow, Spark, PyTorch, Ray, Flink, etc.).
Provides multi-level queue management capabilities, enabling fine-grained resource quota control and task priority scheduling.
Efficiently schedules heterogeneous devices like GPU and NPU, fully unleashing hardware computing potential.
Greatly enhancing model training efficiency in AI distributed training scenarios.
Supports cross cluster job scheduling, improving resource pool management capabilities and achieving large scale load balancing.
Enables online and offline workloads colocation, improving cluster resource utilization through intelligent scheduling strategies.
Optimizing cluster load distribution and enhancing system stability.
Supports various scheduling strategies such as Gang scheduling, Fair-Share, Binpack, DeviceShare, NUMA-aware scheduling, Task Topology, etc.