Profiling Real-Time Applications with Intel® VTune™ Profiler (Intel® Time Coordinated Computing Tools 2021.2)
This tutorial focuses on hotspots related to memory access. In this
context, hotspots are memory objects (such as arrays) in your real-time
application that have a high number of Level 2 (L2) or Level 3 (L3)
Intel® Time Coordinated Computing Tools (Intel® TCC Tools) offers the
cache allocation library
, a set of C APIs that can decrease the number of misses by allocating buffers in L2 and L3 cache. Since cache is smaller than DRAM, it is important to choose carefully which memory objects to target so as not to overuse the cache. Hotspots are the prime candidates. The first step to using the cache allocation library is to find the hotspots in your application.
This tutorial shows how to find hotspots at the L2 and L3 cache levels
with VTune™ Profiler. In general, consider the size and time
sensitivity of the data to decide which cache level to analyze. If the
data size is small and access is highly time critical, it’s more
relevant to check L2 misses. In other cases, check L3 misses.
This tutorial uses a provided sample application as an example.