How is a graphic processing unit (GPU) different from a central processing unit (CPU)?

Question

Options

• AA GPU has a larger cache
• BA GPU has fewer cores.
• CA GPU is more highly parallel.
• DA GPU is task-oriented.

• A
  3% (1)
• B
  3% (1)
• C
  95% (38)

Explanation

GPUs are designed for massive parallelism - they contain thousands of smaller cores optimized to handle many simple operations simultaneously, which is why option C is correct. This architecture makes them ideal for rendering graphics or training neural networks, where the same operation must be applied to enormous datasets at once.

Why the distractors are wrong:

• (A) Larger cache - CPUs typically have larger, more sophisticated cache hierarchies; GPU cache is smaller and optimized for throughput, not latency.
• (B) Fewer cores - The opposite is true; GPUs have far more cores than CPUs (thousands vs. tens).
• (D) Task-oriented - CPUs are the task-oriented processors, designed to handle complex, sequential, general-purpose logic quickly; GPUs are data-oriented.

Memory tip: Think "GPU = Graphics = Grid of pixels processed all at once" - parallelism is baked into their origin. If a question mentions parallel, throughput, or simultaneous computation, GPU is the answer.

No community discussion yet for this question.