Amazon EC2 instance types

Correct Answer: P3 instances, which are optimized for machine learning and high performance computing using NVIDIA Tesla V100 GPUs.

P3 instances are the best choice for running GPU-intensive applications for machine learning tasks. These instances are optimized specifically for machine learning and high-performance computing workloads, and they come equipped with NVIDIA Tesla V100 GPUs, which are powerful and well-suited for these types of tasks. The other instance types mentioned, such as C5 (compute-optimized), R5 (memory-optimized), and M5 (general-purpose), do not have the specialized GPU hardware that is required for efficient machine learning computations. While they may be suitable for other types of workloads, they are not the optimal choice for GPU-intensive machine learning applications.

Correct Answer: I3 instances, which are optimized for high random I/O performance and low latency

For a data-intensive application that requires high disk throughput and low latency, the optimal choice would be I3 instances. I3 instances are specifically designed and optimized for workloads that require high random I/O performance and low latency. They are equipped with NVMe SSD storage, which provides extremely fast and low-latency access to data. This makes them well-suited for applications that heavily rely on disk I/O operations. C5 instances are compute-optimized instances that offer high CPU performance but may not be the best fit for disk-intensive workloads. R5 instances are memory-optimized instances suitable for memory-intensive applications but may not provide the optimal disk performance required in this scenario. T3 instances provide burstable performance and are suitable for workloads with moderate CPU usage, but they are not specifically optimized for high disk throughput and low latency.

Correct Answer: G4

G4 instances are specifically designed for graphics-intensive applications like 3D rendering and provide the latest generation NVIDIA GPUs. The other instance types do not offer the required GPU performance for this type of workload.