Given a trace of my customized workload following the memory trace driven format, can I get a per access latency? My output currently using the LoadStoreTrace frontend just shows per channel stats. Is it possible to get latency numbers for each request somehow?