The SQL Server CPU page shows CPU-related monitoring data for a SQL Server instance. Use it to understand CPU pressure, workload spikes, and whether query tuning or host investigation is needed.
CPU pressure can come from a few expensive queries, many small queries, parallelism, compilation, inefficient plans, or work outside SQL Server on the host. Start by matching the CPU spike to a time window. Then use SQL Server Activity to see live workload and Cached SQL or SQL Server Query Store to find statements with high CPU cost.
Do not assume that more CPU is the only fix. Query tuning, better indexes, plan regression analysis, parameter sensitivity review, and application batching can reduce CPU without changing hardware. If host OS metrics are available, also check whether another process is consuming CPU on the server.
CPU should be compared with waits. If waits show storage or locking as the dominant delay, CPU may be a symptom rather than the bottleneck.
Start with the highest total business impact. A frequent medium-cost query can matter more than a rare expensive one.
Yes. Missing or inefficient indexes can force scans, extra joins, and expensive plan choices.
Configure Host OS Login where appropriate so Mini DBA can collect host-level metrics. For Windows SQL Server hosts, use Enable WMI For SQL Server Monitoring to allow CPU core and process visibility.