When my computer spiked to 90% CPU usage out of the blue, I stared at the Task Manager’s Processes tab for a while and terminated a few processes. I hoped it would resolve the problem, but nothing changed. The Task Manager tends to hide the real triggers for certain problems, so I switched to Event Viewer. Within minutes, I found the process and error that were hammering my processor. Once I’d dealt with them, CPU usage dropped and stayed normal. I’ve stopped using the Task Manager as much ever since.

Navigating the Event Viewer

When the deeper tool is actually faster

Windows keeps a record of everything that happens on a computer, including crashes, warnings, failed updates, driver issues, and app installations. These events can be found in the Event Viewer. However, there are several logs, and it may be confusing to know where to start. System and Application logs are the most likely places to catch noisy background activities that can spike the CPU.

In my case, the repeated errors were in the WMI-Activity log. After navigating to Applications and Services Logs -> Microsoft -> Windows -> WMI-Activity -> Operational, I used the built-in Filter Current Log feature for more details. I set the Event level to Error and Warning, and narrowed the time range (Logged) to the exact window when the spike happened. This cut off the noise, and all I needed to do was look for repeated Event IDs or a source that repeats in a short timeframe. I found the same WMI-Activity errors appearing every few seconds. When I cross-checked against the Task Manager’s Details tab, I found a discrepancy: Event Viewer usually names the actual component (repeated errors from WmiPrvSE.exe in the WMI-Activity log), whereas Task Manager may only show a generic host process.

The Event Viewer doesn’t open by default to the Applications or System logs, which is why Task Manager feels accessible. After navigating and filtering, the details were far more useful than anything I’d gotten from Task Manager, and I could finally see the process that had triggered the spike.

The error that kept restarting in the background

Task Manager never showed the real pattern

What Event Viewer revealed was a specific error that had entered a tight failure loop. WmiPrvSE.exe was constantly crashing, logging the error, and restarting. Even after the computer reboots, it hits the same problem again and crashes.

This behavior may appear normal in Task Manager. The problematic process only appears briefly, spikes the CPU, and disappears. I tried sorting by CPU, but the high usage was so brief and kept on moving between process IDs and generic names like svchost.exe, RuntimeBroker, and WmiPrvSE, so the pattern wasn’t obvious. Seeing the error source repeated over and over within a short window in the Event Viewer was all the clarity I needed. That constant start-crash-restart cycle was what kept my overall CPU usage sitting at 80–90% for long stretches.

Stopping the loop permanently

A few simple steps that ended the spikes for good

After confirming that WmiPrvSE.exe was stuck in a restart loop, I opened Windows Services, located the Windows Management Instrumentation service, and restarted it. This stopped the heavy, relentless resource usage. Next, I opened the Command Prompt with elevated privileges and ran the command below to make the fix stick:

winmgmt /salvagerepository

I rebooted the system and returned to the Event Viewer error log, and the repeated WMI-Activity had stopped. Looking in Task Manager, CPU usage had also dropped from about 90% to normal. The fans, which had been whirring loudly under load, also quieted down.

The problem was over, but I took two more steps: I checked whether a related task was forcing WMI queries, and the next day I returned to the WMI-Activity log to confirm the error had not returned.

The quicker way to troubleshoot Windows

To be fair, Event Viewer feels a bit dated, especially compared with Task Manager, which Windows has improved over successive builds. But even at that, it digs much deeper than Task Manager into the actual sequence of failures on Windows. It’s the more robust tool if you know where to look.

This experience taught me that the next time I have a CPU spike that Task Manager doesn’t catch, the answer is buried somewhere in Event Viewer. Filtering recent Errors and Warnings almost always takes less time than repeatedly ending processes in Task Manager. Additionally, Event Viewer provides enough context to help prevent the spike from recurring.