← MLS-C01·Amazon Web Services

Question 1,370 of 1,755

Modeling →hardMultiple ChoiceObjective-mapped

Model Parallelism vs Data Parallelism — Choosing the Right Distributed Strategy

This MLS-C01 practice question tests your understanding of modeling. Compare every option against the stated constraints before choosing — the best answer satisfies all requirements, not just the most obvious one. After answering, compare your reasoning against the explanation and wrong-answer breakdown below. Once you have made your selection, read the full explanation to reinforce the concept and understand why each distractor is designed to mislead on exam day.

A team is training a deep learning model on SageMaker using a custom PyTorch container. Training takes 24 hours on a single ml.p3.2xlarge instance. The team wants to reduce training time using distributed training. Which strategy is MOST appropriate?

A
Use data parallelism with Horovod across multiple instances
Data parallelism replicates the model on each instance and splits data across instances, reducing training time by processing more data in parallel. This is the most appropriate strategy for a model that fits on a single GPU and needs faster training.
B
Use model parallelism to split the model across multiple GPUs
Why wrong: Model parallelism splits the model across GPUs, which is useful when the model is too large for one GPU. Since the model already fits on a single GPU, this adds unnecessary complexity without significant time reduction.
C
Use SageMaker Managed Spot Training to reduce cost
Why wrong: SageMaker Managed Spot Training reduces cost, not training time. It does not address the need for faster training.
D
Use SageMaker Automatic Model Tuning to find optimal hyperparameters
Why wrong: SageMaker Automatic Model Tuning optimizes hyperparameters, which may improve model performance but does not directly reduce training time.

Full breakdown with real-world context →

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Use data parallelism with Horovod across multiple instances

The training takes 24 hours on a single ml.p3.2xlarge instance with one GPU. To reduce training time, distributing the workload across multiple instances is effective. Data parallelism, where the model is replicated on each instance and each processes different data batches with gradient averaging, is the standard approach for models that fit on a single GPU. Model parallelism is designed for models too large to fit on one GPU, which is not the case here. Therefore, data parallelism with Horovod across multiple instances is the most appropriate strategy.

Key principle: Answer the scenario, not the keyword: identify the specific constraint before choosing the most familiar-sounding option.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Use data parallelism with Horovod across multiple instances
Why this is correct
Data parallelism replicates the model on each instance and splits data across instances, reducing training time by processing more data in parallel. This is the most appropriate strategy for a model that fits on a single GPU and needs faster training.
Related concept
Read the scenario before looking for a memorised answer.
✗
Use model parallelism to split the model across multiple GPUs
Why it's wrong here
Model parallelism splits the model across GPUs, which is useful when the model is too large for one GPU. Since the model already fits on a single GPU, this adds unnecessary complexity without significant time reduction.
✗
Use SageMaker Managed Spot Training to reduce cost
Why it's wrong here
SageMaker Managed Spot Training reduces cost, not training time. It does not address the need for faster training.
✗
Use SageMaker Automatic Model Tuning to find optimal hyperparameters
Why it's wrong here
SageMaker Automatic Model Tuning optimizes hyperparameters, which may improve model performance but does not directly reduce training time.

Common exam traps

Common exam trap: answer the scenario, not the keyword

The MLS-C01 exam often tests the misconception that model parallelism is the best distributed training strategy, but here the model fits on one GPU and the goal is to reduce time, making data parallelism across instances more suitable.

Detailed technical explanation

How to think about this question

Model parallelism partitions the model layers across multiple GPUs, so each GPU computes a subset of the forward and backward passes, reducing memory pressure and enabling larger batch sizes per GPU. In PyTorch, this can be implemented using `torch.distributed.pipeline.sync.Pipe` or manual layer placement with `nn.Module.to(device)`. For a 24-hour single-GPU job, model parallelism can achieve near-linear speedup if the model has significant sequential depth, but it requires careful balancing of layer sizes to avoid pipeline bubbles.

KKey Concepts to Remember

Read the scenario before looking for a memorised answer.
Find the constraint that changes the correct option.
Eliminate answers that are true in general but not in this case.

TExam Day Tips

Watch for words such as best, first, most likely and least administrative effort.
Review why wrong options are wrong, not only why the correct option is correct.

Key takeaway

Answer the scenario, not the keyword: identify the specific constraint before choosing the most familiar-sounding option.

Real-world example

How this comes up in practice

A startup's cloud architect reviews their monthly bill and notices costs are higher than expected for a long-running batch job. Switching from on-demand instances to Reserved Instances — or using Spot/Preemptible VMs — can reduce compute costs by up to 72 %. Questions like this test whether you understand the tradeoffs between commitment, flexibility, and cost across cloud pricing models.

What to study next

Got this wrong? Here's your next step.

Identify which exam domain this question belongs to, review the core concept, then practise similar questions from the same domain.

Related MLS-C01 practice-question pages

Use these pages to review the topic behind this question. This is how one missed question becomes focused revision.

Data Engineering practice questions

Practise MLS-C01 questions linked to Data Engineering.

Machine Learning Implementation and Operations practice questions

Practise MLS-C01 questions linked to Machine Learning Implementation and Operations.

Modeling practice questions

Practise MLS-C01 questions linked to Modeling.

Exploratory Data Analysis practice questions

Practise MLS-C01 questions linked to Exploratory Data Analysis.

MLS-C01 fundamentals practice questions

Practise MLS-C01 questions linked to MLS-C01 fundamentals.

MLS-C01 scenario practice questions

Practise MLS-C01 questions linked to MLS-C01 scenario.

MLS-C01 troubleshooting practice questions

Practise MLS-C01 questions linked to MLS-C01 troubleshooting.

Practice this exam

Start a free MLS-C01 practice session

Short sessions build daily habit. Longer sessions build exam-day stamina. Try a timed session to simulate real conditions.

10 questions 20 questions 30 questions 50 questions Timed 30

MLS-C01 practice-test guide →Study guide →Browse all practice tests

FAQ

Questions learners often ask

What does this MLS-C01 question test?

Modeling — This question tests Modeling — Read the scenario before looking for a memorised answer..

What is the correct answer to this question?

The correct answer is: Use data parallelism with Horovod across multiple instances — The training takes 24 hours on a single ml.p3.2xlarge instance with one GPU. To reduce training time, distributing the workload across multiple instances is effective. Data parallelism, where the model is replicated on each instance and each processes different data batches with gradient averaging, is the standard approach for models that fit on a single GPU. Model parallelism is designed for models too large to fit on one GPU, which is not the case here. Therefore, data parallelism with Horovod across multiple instances is the most appropriate strategy.

What should I do if I get this MLS-C01 question wrong?

Identify which exam domain this question belongs to, review the core concept, then practise similar questions from the same domain.

What is the key concept behind this question?

Read the scenario before looking for a memorised answer.

About these practice questions

Courseiva creates original exam-style practice questions with explanations and wrong-answer analysis. It does not publish real exam questions, exam dumps, or protected exam content. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

Keep practising

Question Discussion

Share a tip, memory trick, or ask about the reasoning behind this question. Do not post real exam questions, leaked content, braindumps, or copyrighted exam material. Comments are moderated and may be removed without notice.

Loading comments…

This MLS-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLS-C01 exam.