Analyze Memory Performance with Precision An important factor in determining application performance is the time required for the application to fetch data from the processor’s cache hierarchy and from the memory subsystem. In a multi-socket system where Non-Uniform Memory Access (NUMA) is enabled, local memory latencies and cross-socket memory latencies will vary significantly. Besides latency, bandwidth (b/w) also plays a big role in determining performance. So, measuring these latencies and b/w is important to establish a baseline for the system under test, and for performance analysis.

Features

• Latency Measurement: Accurately measure memory latency across NUMA nodes and memory hierarchies.
• Bandwidth Analysis: Evaluate memory bandwidth under different access patterns and workloads.
• Load Sensitivity Testing: Observe how latency and bandwidth change with increasing system load.
• NUMA Awareness: Analyze memory access behavior in non-uniform memory access (NUMA) systems.
• Performance Tuning Aid: Identify bottlenecks and optimize memory usage for high-performance computing.