Q What is the system memory latency ladder?

A The latency ladder shows data access times in ns for different block sizes.

---

The system memory latency ladder shows random access times in billionths of a second (ns) for block sizes that cover each of the various CPU Caches (typicaly L1 = 32KB, L2 = 256KB, L3 = 4MB) and main memory (typically 4MB+). The varying latencies show how quickly a CPU can access data from different locations. Although lower latencies improve system performance they generally have less effect than overall bandwidth which is the primary performance metric for RAM. Latency measures the minimum time required for access to individual blocks of data whereas throughput measures the maximum number blocks that can be delivered per second.