Exam NCA-AIIO Topic 1 Question 98 Discussion

Actual exam question for NVIDIA's NCA-AIIO exam
Question #: 98
Topic #: 1
In a 512-GPU training cluster, engineers observe poor scaling efficiency when moving from single-node to multi-node runs, even though the compute nodes themselves are not fully utilized and the model code has already been optimized. Profiling shows that collective all-reduce operations dominate step time as node count increases. Why should InfiniBand be chosen over traditional Ethernet for this type of multi-node AI training environment?

Suggested Answer: D Vote an answer

InfiniBand provides high-bandwidth, low-latency connectivity with hardware support for collective communication primitives like all-reduce, which are critical for synchronizing gradients across GPUs in multi-node distributed training, directly improving scaling efficiency.

by Viola at Jul 30, 2026, 11:35 AM

Limited Time Offer

15%

Off

Get Premium NCA-AIIO Questions as Interactive Self Test Engine or PDF

Comments

Chosen Answer:
This is a voting comment (?) , you can switch to a simple comment.
Switch to a voting comment New
Nick name: Submit Cancel
A voting comment increases the vote count for the chosen answer by one.

Upvoting a comment with a selected answer will also increase the vote count towards that answer by one. So if you see a comment that you already agree with, you can upvote it instead of posting a new comment.
0
0
0
10