Exam Associate Cloud Engineer topic 1 question 312 discussion

Actual exam question from Google's Associate Cloud Engineer

Question #: 312
Topic #: 1

[All Associate Cloud Engineer Questions]

Your company's machine learning team requires a scalable and flexible platform to fine-tune large language models utilizing a large volume of proprietary data on Google Cloud. You are tasked with building a solution for this team. What should you do?

A. Use Dataflow as a platform to run the fine-tuning jobs
B. Use a Compute Engine managed instance group as a platform to deploy Jupyter Notebooks and run fine-tuning jobs.
C. Use Cloud Run and GPU as a platform to run the fine-tuning jobs.
D. Use Google Kubernetes Engine (GKE) and hardware accelerators as a platform to run the fine-tuning jobs.

Show Suggested Answer

Suggested Answer: D 🗳️

by toasty at July 6, 2025, 7:33 a.m.

Comments

Submit Cancel

toasty

1 month ago

Selected Answer: D

GKE with hardware accelerators provides the robust, scalable, and flexible platform required by a machine learning team for fine-tuning large language models. It is the industry-standard for orchestrating containerized machine learning workloads at scale, offering the necessary control and automation for complex, distributed training jobs. The other options are either not designed for this purpose (Dataflow) or lack the comprehensive orchestration and scalability features of GKE for this specific use case (Compute Engine MIGs and Cloud Run).

upvoted 1 times

...