What enables moving data between GPU memory and local or remote storage without using the CPU?

The correct answer is D. GPUDirect Storage. NVIDIA GPUDirect Storage enables direct data paths between GPU memory and local or remote storage (e.g., NVMe over fabrics), bypassing the CPU and host memory. This maximizes throughput and minimizes latency in AI data pipelines. NVLink connects GPUs, GPUDirect P2P facilitates…

NVIDIA AI Platform and Ecosystem

Question

Options

ANVLink
BGPUDirect P2P
CInfiniBand
DGPUDirect Storage

How the community answered

(34 responses)

A
3% (1)
B
6% (2)
C
3% (1)
D
88% (30)

Explanation

NVIDIA GPUDirect Storage enables direct data paths between GPU memory and local or remote storage (e.g., NVMe over fabrics), bypassing the CPU and host memory. This maximizes throughput and minimizes latency in AI data pipelines. NVLink connects GPUs, GPUDirect P2P facilitates GPU-to- GPU transfers, and InfiniBand is a network fabric, but only GPUDirect Storage targets storage access.

Topics

#GPUDirect Storage#GPU data transfer#Zero-copy I/O#Storage acceleration

Community Discussion

No community discussion yet for this question.

Full NCA-AIIO Practice