Answer
Check how many rows exist in ifsiamsys.EVENT_ENTITY and ifsapp.IAM_LOGIN_EVENT_DETAIL, and specifically how many are older than 30 days — IAM is meant to retain only the last 30 days of login-event data in these tables, and a known cleanup gap can leave them growing unbounded, which drives up CPU as IAM repeatedly scans over the bloated tables. Clearing out the stale rows (per IFS's guidance for this specific known issue) is the fix, rather than treating it as a general capacity problem.