Accelerated Data Models are the engine of ES. They build high-performance TSIDX (Time-Series Index) summaries of CIM-mapped data.
By using `tstats` on these accelerated models, Splunk avoids reading raw data from disk, making massive SIEM searches exponentially faster.