Start with the process and the task it is doing. Repeated samples are more useful than a single high reading.
By MacMonitor · Published · Updated · macOS 26; app requires macOS 14 or later
Short answer: open Activity Monitor → CPU, choose View → All Processes, and sort by % CPU. Compare several samples while reproducing the slowdown. A short burst during a build or export is expected; persistent activity after the task ends needs investigation.
A busy process can use more than 100% across multiple CPU cores. Compare like-for-like measurements; a per-process figure is not necessarily on the same scale as an overall CPU chart. CPU Time is cumulative time, not the current load.
Indexing, backups, builds and video exports are examples of work that can keep processors busy. Let a known task finish when practical. If responsiveness is poor with sustained memory pressure, investigate memory as well as CPU.
“My Mac is slow” is a starting observation, but it is too broad to identify a cause. Try to name the specific task: opening a project, joining a call, importing photographs, or leaving a browser idle. Then watch the CPU view while that task runs. You want to connect a process with an activity, not identify a suspicious-looking name at random.
Write down whether the CPU activity stops when the job completes. A build that is busy while compiling is different from a process that remains busy long after its window closes. Save useful work before experimenting, and keep each change small enough that you can explain what happened afterward.
Activity Monitor can display individual processes alongside a system-wide summary. Those views have different scales and purposes. A process working across several cores can display a figure above 100%, while an overall chart summarizes the machine differently. Read the label and time window before comparing a number with another tool.
CPU Time is also different from % CPU. A long-running process may have accumulated substantial CPU Time without being the process that is busy now. Sort the current usage column for the immediate investigation, and use cumulative measures only when their meaning fits your question.
Imagine a video export beside a browser extension that repeatedly processes a page. This is a hypothetical example. Both can appear near the top of the CPU list, but the export has an expected purpose and endpoint. The extension needs closer investigation if it remains busy after the page or task should have settled.
Let the known export finish when practical. Then compare the browser with the relevant page or extension disabled through its own controls. If the activity stops, repeat the normal task to see whether the pattern returns. This is more informative than force-quitting both applications and attributing the improvement to a generic cleanup.
A slowdown can involve several constraints at once. If pressure remains high in the Memory view, follow the memory-pressure procedure rather than assuming the busiest process is using CPU inefficiently. If a task cannot write its output because the disk is full, inspect available storage separately.
Keep those investigations distinct. Deleting caches is not a general repair for a CPU-heavy task, and a CPU graph alone does not prove that more memory or a larger disk is needed. Use the task, the relevant measurements and the effect of one deliberate change together.
An ordinary application may offer to save work when you quit through its own menu. Force Quit can interrupt that opportunity. If the application is responsive, save your documents and close it normally before considering a stronger action. If it is unresponsive, consider what unsaved work could be lost before proceeding.
Do not use Force Quit on unfamiliar system processes simply to reduce the graph. Some processes are performing work on behalf of other apps or macOS. Identify the owner and consult its documentation when you cannot explain the activity. A process name alone does not establish malware or justify deleting its files.
A restart can clear a transient state and allow normal work to resume. It can also erase the context needed to understand the original event. If the Mac is usable enough, note the process and task before restarting. Do not claim that the underlying cause is fixed solely because the graph is quiet immediately afterward.
If the same sequence produces the same sustained activity again, collect a focused report for the relevant developer. Include app and macOS versions, the actions taken, whether a plugin was enabled, and the difference after a controlled change. Remove private filenames and content before sharing it.
| Record | Example of useful detail |
|---|---|
| Task | Exporting a particular project or opening a specific kind of document. |
| Process and owner | The application or extension associated with the busy row. |
| Duration | Whether activity stops with the task or continues afterward. |
| One change | A normal app restart, one disabled extension, or a completed background task. |
| Result | Whether the original work becomes responsive under comparable conditions. |
No. A demanding, understood task can use substantial processing capacity. Interpret the scale, duration and effect on your work. The useful question is whether the activity is expected and whether it ends when the task ends.
Yes, applications can use helper processes for separate tasks. Compare the application and its helpers as a workload, and use its own controls to close a tab, job or extension when possible. A grouped monitor can help summarize that view, while Activity Monitor provides individual process detail.
A cleaner is not a general CPU repair tool. It may help review storage when storage is part of the problem, but persistent processor activity needs a task-specific investigation. Avoid promises of automatic optimization or a fixed percentage speed increase.
If sustained unexplained activity repeatedly disrupts normal work, provide the relevant developer or Apple Support with the reproducible sequence and observations. A concise record is more useful than repeated screenshots without context.
Review the owner and purpose of an item before changing its behavior. A backup, sync service or scheduled development job may be doing useful work. Disable only an understood optional component through supported controls, then compare the original task again. Changing many startup and background settings simultaneously can make a quiet graph harder to interpret, because you no longer know which service affected the result or which useful function stopped running.
MacMonitor Activity shows live CPU and available app readings, plus optional sustained-load alerts. Its local-server Stop action is confirmed and limited to eligible developer processes. It does not automatically optimize or repair an app.