Choosing the right cache replacement policy is critical for maintaining a high cache hit ratio and, therefore, low-latency access, but the decision can be workload-specific