Answer: For GPU clusters, it's best to configure parallelism to match the number of GPU-enabled instances, preventing task conflicts.

The correct answer is **B) Configure parallelism based on the number of GPU-enabled instances, avoiding conflicts among Spark tasks.** Key considerations include: 1. **Executor Thread Configuration:** - **CPU vs. GPU:** CPU clusters often use multiple executor threads per node to maximize CPU utilization. In contrast, GPU clusters typically use one executor thread per node to prevent conflicts among tasks accessing the same GPU, which is optimal for GPU-accelerated libraries. 2. **Parallelism Adjustment:** - **Reduced Maximum Parallelism:** The single-executor-thread-per-node setup on GPU clusters means lower maximum parallelism compared to CPU clusters. - **Manual Configuration:** Parallelism should be manually set to match the number of GPU-enabled instances in your cluster to efficiently use resources and avoid conflicts. **How to configure parallelism:** - **Determine the Number of GPUs:** Identify the total GPU-enabled instances in your cluster. - **Adjust Spark Configuration:** Set Spark options like `spark.executor.instances`, `spark.executor.cores`, and `spark.default.parallelism` according to the number of GPUs. **Example Configuration:** ```python spark = SparkSession.builder.appName("my_app").config( spark.executor.instances=4, spark.executor.cores=1, spark.default.parallelism=4 ).getOrCreate() ``` **Additional Tips:** - Choose GPU instance types that meet your workload's needs. - Ensure your libraries support GPU acceleration. - Monitor GPU usage to spot and address bottlenecks.

Author: LeetQuiz Editorial Team