Which workload is best suited for using TPUs instead of GPUs?

Large-scale neural network training

General-purpose scientific computing

A company has a monolithic application that is difficult to containerize. They need to migrate it to Google Cloud with minimal changes and want to reduce operational overhead. The application has variable usage patterns. Which compute option should they choose?

Compute Engine with a managed instance group and autoscaling

A team has built a container image and pushed it to Google Artifact Registry. They want to deploy it to an existing GKE cluster. What is the first step they should take?

Use kubectl to create a deployment that references the container image.

Use gcloud to create a new cluster.

Use Cloud Build to build the image.

Use Docker to run the container locally.

A data analytics team needs to run a memory-intensive job that requires 128 GB of RAM and 16 vCPUs. The predefined machine types do not offer this exact combination. What should they do?

Use a predefined machine type with 16 vCPUs and 128 GB of memory.

Use a Spot VM with a predefined machine type.

Create a custom machine type with 16 vCPUs and 128 GB of memory.

What is a common use case for Cloud Run functions?

Running machine learning training jobs

Hosting a long-running web server

Google CloudAssociate Cloud Engineer

Domain 2Objective 1

2.1 Planning and Implementing Compute Resources ASSOCIATE-CLOUD-ENGINEER Practice Questions (Page 4)

Part of the Planning and implementing a cloud solution domain, which accounts for ~30% of the ASSOCIATE-CLOUD-ENGINEER exam.

36questions here

8free pages

12concepts

~30%of the exam

12 Concepts in This Objective

Compute options selection
Compute Engine instance provisioning
Compute Engine storage selection
Managed instance group autoscaling
OS Login configuration
VM Manager configuration
Spot VMs and custom machine types
kubectl installation and configuration
GKE cluster deployment configurations
Containerized application deployment to GKE
Serverless application deployment
GPU and TPU selection

Questions 16–20

16foundation · easy
Which workload is best suited for using TPUs instead of GPUs?
Select an answer first
17expert · hard
A company has a monolithic application that is difficult to containerize. They need to migrate it to Google Cloud with minimal changes and want to reduce operational overhead. The application has variable usage patterns. Which compute option should they choose?
Select an answer first
18application · medium
A team has built a container image and pushed it to Google Artifact Registry. They want to deploy it to an existing GKE cluster. What is the first step they should take?
Select an answer first
19application · medium
A data analytics team needs to run a memory-intensive job that requires 128 GB of RAM and 16 vCPUs. The predefined machine types do not offer this exact combination. What should they do?
Select an answer first
20foundation · easy
What is a common use case for Cloud Run functions?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to ASSOCIATE-CLOUD-ENGINEER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Google Cloud. “ASSOCIATE-CLOUD-ENGINEER” is a trademark of its owner, used for identification only.