Select a node type with a higher number of CPU cores and enable autoscaling to dynamically adjust resources, ensuring both memory and CPU are optimized for varying workloads.

Choose a node type with a larger amount of memory per node, as this directly addresses the memory requirements of the workload without unnecessarily increasing costs with additional CPU cores.

Configure the cluster to use a custom disk type with higher I/O performance, under the assumption that disk speed will compensate for memory limitations.

Implement a combination of node types, using some with high memory for memory-intensive tasks and others with high CPU for processing tasks, to balance cost and performance.