The LLC (last-level cache) is the last, and longest-latency, level in the memory hierarchy before main memory (DRAM). While LLC hits are serviced much more quickly than hits in DRAM, they can still incur a significant performance penalty. This metric also includes coherence penalties for shared data.
A significant proportion of cycles is being spent on data fetches that miss in the L2 but hit in the LLC. This metric includes coherence penalties for shared data.
1. If contested accesses or data sharing are indicated as likely issues, address them first. Otherwise, consider the performance tuning applicable to an LLC-missing workload: reduce the data working set size, improve data access locality, consider blocking or partitioning your working set so that it fits into the low-level cache, or better exploit hardware prefetchers.
2. Consider using software prefetchers, but note that they can interfere with normal loads, potentially increasing latency, as well as increase pressure on the memory system.