NVIDIA driver update leaves laptop GPU busy: 100% fan, high battery drain and persistent nvcontainer GPU usage

Environment

Windows 11 Pro 22H2 laptop with hybrid graphics (Intel iGPU + NVIDIA RTX 3050 mobile), NVIDIA driver 531.41 installed via GeForce Experience, OEM power profile default, common apps: Chrome, VS Code, Slack.

Symptoms

After installing the new NVIDIA driver, fans run at max RPM constantly, battery life drops from ~6 hours to ~1.5 hours, and Task Manager shows high GPU engine usage despite no obvious GPU workloads. Graphics-intensive games run fine but laptop remains loud during light web browsing.

Observations

nvidia-smi shows persistent GPU process 'NvContainer.exe' and 'NvContainerLocalSystem.exe' with non-zero GPU % and GPU memory pinned. Windows Event Viewer shows repeated entries from 'Display' with warnings about 'nvlddmkm' recovery events. Disabling NVIDIA adapter in Device Manager reduces fan noise but causes external monitor to lose output until reboot. Clean install using Device Manager -> uninstall driver did not remove NvContainer services (they reappear after reinstall).

Already attempted

Rolled back driver to previous version using Device Manager (reversion appeared to partially work but after reboot nvcontainer reappeared). Used DDU in Safe Mode to remove all NVIDIA drivers, then installed older 526.xx driver — machines still reported high fan speeds after install; created a new Windows power plan and set 'Graphics preference' to integrated for Chrome/Edge, which reduced but did not eliminate the symptom. Performed malware scan (clean).

Question

Why does the NVIDIA driver update cause persistent GPU usage and fan/battery issues even when no heavy apps are running, and how can I fully revert to a safe previous state without losing system stability?

Post timeline

COMMUNITY RESPONSES

Verified community troubleshooting

0 replies

No answers yet..

Showing 0 of 0 community responses

AI DIAGNOSTICS

Incident analysis

Updated 1w ago• Trouble ID

Most likely causes

New NVIDIA driver contains a bug that prevents proper GPU power state transitions, leaving GPU in P0 or high-performance state. 90%

Driver/firmware incompatibility with OEM power management or hybrid graphics switching (Optimus) implementation. 85%

NvContainer or telemetry service actively uses the GPU for hardware acceleration (e.g., NVENC/decoding) due to telemetry or background tasks. 80%

Summary

Persistent GPU usage by nvcontainer and display driver warnings after update indicate the driver or its system services are holding the GPU in a high-power state. Root causes include driver bugs (resource leak or mis-set power-management state), background NVIDIA telemetry or helper services incorrectly querying/locking the GPU, or incompatibility between OEM firmware and new driver power-management defaults. Device Manager 'disable' shows the display stack depends on NVIDIA drivers for external monitor routing, explaining continued issues after partial removal. Full, clean removal using DDU normally reverts behavior; however, stubborn services or Windows update driver reinstallation can reintroduce problematic driver.

Suggested Fixes

Diagnostic step 1

Collect diagnostics before making further changes: run nvidia-smi -q (from an elevated PowerShell with NVIDIA tools if available) and capture Event Viewer logs (Display and System) covering the timeframe when symptoms started.

Diagnostic step 2

Fully remove NVIDIA drivers using Display Driver Uninstaller (DDU) in Safe Mode, then reboot to Windows Update 'Optional updates' and explicitly hide NVIDIA driver updates (to prevent automatic reinstallation) until you have a working version.

Diagnostic step 3

Install a known-good older driver from NVIDIA download archive (not GeForce Experience auto-installer). Prefer OEM-recommended drivers if the machine vendor provides a tested package.