When designing a solution on ECS, which of the following node types is best suited for a high- throughput data processing workflow?

Question

Options

• AGeneral-purpose compute node
• BHigh Memory node
• CGPU-accelerated compute node
• DStorage-optimized node

• A
  7% (3)
• B
  12% (5)
• C
  76% (31)
• D
  5% (2)

Explanation

GPU-accelerated compute nodes are purpose-built for massive parallelism - their thousands of cores can process enormous data volumes simultaneously, which is the defining trait of high-throughput workflows like ML inference, real-time analytics, and large-scale data transformation.

Why the distractors fall short:

• A (General-purpose): Balanced CPU/memory/network, but lacks the parallel processing density needed to sustain high throughput at scale.
• B (High Memory): Optimized for memory-resident workloads (e.g., in-memory databases), not raw processing throughput - memory capacity ≠ processing speed.
• D (Storage-optimized): Maximizes disk I/O and capacity, which helps with data access but not data processing speed.

Memory tip: Think of it this way - GPU = Gear for parallel processing. When the question says "throughput," look for the node type that can do the most simultaneous work, not just the most storage or the most RAM.

No community discussion yet for this question.