Compare and contrast GPU and CPU architectures: Practice Questions — Essential AI Knowledge (NVIDIA-Certified Associate: AI Infrastructure and Operations)

Practice Questions: Compare and Contrast GPU and CPU Architectures These multiple-choice questions are designed to help you prepare for the...

Practice Questions: Compare and Contrast GPU and CPU Architectures

These multiple-choice questions are designed to help you prepare for the NVIDIA-Certified Associate: AI Infrastructure and Operations exam by testing your understanding of the differences and similarities between GPU and CPU architectures.

Which of the following best describes the primary architectural difference between a GPU and a CPU?
- A. CPUs have many cores optimized for parallel processing, GPUs have few cores optimized for sequential tasks.
- B. CPUs have fewer cores optimized for sequential processing, GPUs have many cores optimized for parallel processing.
- C. CPUs and GPUs have the same number of cores but differ in clock speed.
- D. CPUs are designed only for graphics rendering, GPUs are designed only for general computing.
Correct Answer: B
Explanation: CPUs typically have fewer cores optimized for sequential task execution, while GPUs have many smaller cores designed to handle thousands of parallel threads simultaneously, making them ideal for AI workloads.
In AI training workloads, why are GPUs generally preferred over CPUs?
- A. GPUs consume less power than CPUs.
- B. GPUs have higher clock speeds than CPUs.
- C. GPUs can perform massive parallel computations efficiently, accelerating matrix operations.
- D. CPUs have better support for floating-point operations.
Correct Answer: C
Explanation: AI training involves large-scale matrix and tensor operations that benefit from the parallel architecture of GPUs, enabling faster computation compared to CPUs.
Which statement accurately reflects the memory architecture difference impacting AI workloads?
- A. CPUs have higher memory bandwidth than GPUs.
- B. GPUs have specialized high-bandwidth memory designed to feed many cores simultaneously.
- C. CPUs and GPUs use the same type of memory with identical bandwidth.
- D. GPUs have less memory capacity than CPUs, limiting their AI capabilities.
Correct Answer: B
Explanation: GPUs utilize high-bandwidth memory (e.g., HBM) to supply data to thousands of cores efficiently, which is critical for AI computations, whereas CPUs generally have lower memory bandwidth.
Which of the following is a typical use case where CPUs outperform GPUs?
- A. Large-scale deep learning model training.
- B. Real-time graphics rendering.
- C. Complex sequential logic and control flow tasks.
- D. Massive parallel matrix multiplication.
Correct Answer: C
Explanation: CPUs excel at executing complex sequential instructions and control flow due to their sophisticated cores and cache hierarchy, whereas GPUs are optimized for parallel workloads.
How does the instruction set architecture (ISA) differ between CPUs and GPUs in the context of AI workloads?
- A. CPUs use a fixed instruction set; GPUs use a programmable instruction set optimized for parallelism.
- B. CPUs and GPUs share the same ISA.
- C. GPUs have no instruction set and rely on CPUs for instructions.
- D. CPUs use a programmable instruction set; GPUs use a fixed instruction set.
Correct Answer: A
Explanation: CPUs have a fixed and complex instruction set optimized for general-purpose tasks, while GPUs use a programmable instruction set designed to efficiently handle parallel workloads typical in AI.
Which factor contributes most to the rapid adoption of GPUs in AI infrastructure?
- A. Lower cost compared to CPUs.
- B. Ability to accelerate both AI training and inference through parallelism.
- C. Smaller physical size than CPUs.
- D. GPUs consume less power than CPUs.
Correct Answer: B
Explanation: GPUs accelerate AI workloads by enabling massive parallelism, which significantly speeds up both training and inference phases, driving their widespread adoption in AI infrastructure.
In the AI deployment lifecycle, why might CPUs still be used alongside GPUs?
- A. CPUs handle data preprocessing and orchestration tasks that require sequential processing.
- B. CPUs are faster at matrix multiplication than GPUs.
- C. GPUs cannot run AI inference tasks.
- D. CPUs consume more power and thus improve performance.
Correct Answer: A
Explanation: CPUs are often used for tasks requiring sequential processing, such as data preprocessing, system orchestration, and running control logic, complementing GPUs that handle parallel AI computations.

More in this topic

Ready to test your knowledge?

Put what you've learned into practice with a quick quiz and track your progress.

Test your knowledge →