What is a direct benefit of using GPUDirect RDMA for multi-server workloads?

• A.Raises GPU base memory clock speeds.
• B.Offloads data movement from CPUs.
• C.Allows CPUs to prioritize scheduling.
• D.Compresses transferred data.

Comment: *

Name: *

Email: *

Verification: *

Which architecture, training or inference, requires more data storage?

• A.Inference architecture requires more data storage.
• B.Training architecture requires more data storage.
• C.Training and inference architecture require the same amount of data storage.

Comment: *

Name: *

Email: *

Verification: *

You are managing an AI infrastructure where multiple AI workloads are being run in parallel, including image recognition, natural language processing (NLP), and reinforcement learning. Due to limited resources, you need to prioritize these workloads. Which AI workload should you prioritize first to ensure the best overall system performance and resource allocation?

• A.Image recognition
• B.Reinforcement learning
• C.Natural Language Processing (NLP)
• D.Background data preprocessing

Comment: *

Name: *

Email: *

Verification: *

Which library removes the need for developers to optimize their applications for specific machines?

• A.RoCE
• B.MIG
• C.NCCL

Comment: *

Name: *

Email: *

Verification: *

Your AI infrastructure team is managing a deep learning model training pipeline that uses NVIDIA GPUs.
During the model training phase, you observe inconsistent performance, with some GPUs underutilized while others are at full capacity. What is the most effective strategy to optimize GPU utilization across the training cluster?

• A.Reconfigure the model to use mixed precision training.
• B.Reduce the number of GPUs assigned to the training task.
• C.Use NVIDIA's Multi-Instance GPU (MIG) feature to partition GPUs.
• D.Turn off GPU auto-scaling to prevent dynamic resource allocation.

Comment: *

Name: *

Email: *

Verification: *