1. Proposed title of this feature request

KAI scheduler

2. What is the nature and description of the request?

Enable KAI scheduler 

git repo:https://github.com/NVIDIA/KAI-Scheduler

Key Features We Need:

• Fractional GPU support without explicit GPU operator configuration
• Advanced scheduling capabilities (fair scheduling, bin packing)
• Better GPU resource management compared to Kubernetes default scheduler

Goal: boost GPU utilization, reduce fragmentation/idle time, and ensure fair/resource-efficient scheduling for LLM training/inference.

3. Why does the customer need this? (List the business requirements here)

Current Gap:

• We understand OpenShift AI now ships with Kueue support, but Kueue relies on the -Kubernetes default scheduler and lacks the GPU-specific features that KAI provides.

4. List any affected packages or components.

Kueue