Exam Associate Cloud Engineer topic 1 question 81 discussion

D is the correct answer. https://cloud.google.com/kubernetes-engine/docs/how-to/gpus

D. Add a new, GPU-enabled, node pool to the GKE cluster. Ask your ML team to add the cloud.google.com/gke -accelerator: nvidia-tesla-p100 nodeSelector to their pod specification.

The correct answer is D

D is the correct answer.

Option D is good

Creating new node pool w/ GPU-enabled instances is cost - saving solution. This way ML team workload will GPU instance for their ML and other team workload will run smoothly

D makes more cost effective

Answer D. Add a new, GPU-enabled, node pool to the GKE cluster. Ask your ML team to add the cloud.google.com/gke-accelerator: nvidia-tesla-p100 nodeSelector to their pod specification. Adding a new node pool with GPUs is the best option because it allows for a separate set of nodes that can be specifically allocated to workloads that require GPU acceleration, such as the Machine Learning (ML) team's workloads. This approach will not affect other workloads running on the original nodes, keeping the costs low and the overall cluster performance stable.

D is the correct answer.

D is correct

D is correct

D is correct Because if you create entirely new node pool then its not cost efficient and also the pods which not require that much high GPU is get scheduled into it. So instead of that add a new node pool with GPU and in the pod YAML file mention the node affinity to get scheduled into the GPU enabled node pool.

By looking at all answers first 3 can be eliminated without any second thought. D is correct.

Go for D

In this case it is about adding a GPU enabled node pool not a GPU to an existing node-pool

D is the correct answer

D would be the right option when there is possibility to add GPUs without recreating the nodes.