Support quota for core and memory

[ZDWWWWW]

1. Issue or feature description

resourceQuota or volcano-queue can not limit core and memory between different users.
for example:

apiVersion: v1
kind: Pod
metadata:
  name: gpu-pod
spec:
  containers:
    - name: ubuntu-container
      image: ubuntu:18.04
      command: ["bash", "-c", "sleep 86400"]
      resources:
        limits:
          nvidia.com/gpu: 2 # requesting 2 GPUs
          nvidia.com/gpumem: 1000 # Each vGPU contains 1000m device memory （Optional,Integer）
          nvidia.com/gpucores: 10 # Each vGPU uses 10% of the entire GPU （Optional,Integer)

Since the usage of gpucores and gpumem is related to the number of GPUs used, for example, gpu:2, gpucores:10 means that the two cards use 10core respectively, and the total usage is 20core.

resourceQuota or volcano-queue can not limit core and memory between different users.

1. Core and mem are not registered on the node, and resourcequota or volcano-queue cannot sense this resource.
2. Since the usage of gpucores and gpumem is related to the usage of gpu, core and mem cannot be deducted correctly.

[github-actions]

Hi @ZDWWWWW,
Thanks for opening an issue!
We will look into it as soon as possible.

Details

Instructions for interacting with me using comments are available here.
If you have questions or suggestions related to my behavior, please file an issue against the gh-ci-bot repository.