RTX 4050 Laptop stuck in P3 / low power cap when AC is plugged in during GPU load; restarting nvidia-powerd fixes it

System:

  • Laptop: Acer laptop with Intel iGPU + NVIDIA GeForce RTX 4050 Laptop GPU
  • OS: Arch Linux
  • Desktop: Hyprland / Wayland
  • External monitor connected to the NVIDIA dGPU, so the dGPU remains active
  • NVIDIA driver: 610.43.02
  • NVIDIA packages: nvidia-open-dkms / nvidia-utils 610.43.02
  • Service involved: nvidia-powerd
  • Platform profile: balanced-performance

Problem:
The NVIDIA GPU can get stuck in a low performance state after an AC adapter
transition happens while the GPU is under 3D load.

The GPU remains in P3 with low effective power draw, around 30-45W, and
clocks_event_reasons.sw_power_cap=Active, even after AC power is restored.
Temperatures are low and there is no thermal slowdown.

How to reproduce:
Method 1:

  1. Start with AC connected.
  2. Run a GPU stress test or any relevant 3D workload.
  3. While the workload is running, unplug AC.
  4. Plug AC back in while the same workload is still running, or while another GPU
    workload is running.
  5. The GPU may remain stuck in P3 / low power instead of returning to normal AC
    performance.

Method 2:

  1. Start on battery.
  2. Run a GPU stress test or any relevant 3D workload.
  3. Plug AC in while the workload is running.
  4. The GPU may remain stuck in P3 / low power.

Expected behavior:
After AC is plugged in, the GPU should return to normal AC performance, P0 and around
70-75W sustained power draw on this laptop.

Actual behavior:
The GPU remains stuck in P3 with low power draw, usually around 30W, with
sw_power_cap=Active.

Important observations:

  • If AC is plugged in while the GPU has very low usage / is idle, the bug is
    corrected or does not happen.
  • Restarting nvidia-powerd also fixes the issue immediately, even while the stress
    test is still running:
    sudo systemctl restart nvidia-powerd
  • This suggests the stuck state is related to nvidia-powerd / Dynamic Boost / AC

Evidence from attached CSV:

  • Normal run:
    15:20:50 → 15:21:53

  • Bug triggered:
    AC changed from 0 to 1 at 15:23:09 while GPU load was about 98%.
    After that, GPU stayed in P3, about 29-31W, sw_power_cap=Active, with low
    temperature.

  • Recovery:
    AC changed from 0 to 1 at 15:27:02 while GPU was idle.
    Next stress run returned to P0 and about 73W average.

  • Bug triggered again:
    AC changed from 0 to 1 at 15:30:19 while GPU load was 100%.
    GPU stayed in P3 with sw_power_cap=Active.

This does not look thermal:

  • hw_thermal_slowdown: Not Active
  • hw_power_brake_slowdown: Not Active
  • applications_clocks_setting: Not Active
  • temperatures were low, around 47-50C during the stuck state.

Workaround:
Restarting nvidia-powerd restores performance:
sudo systemctl restart nvidia-powerd

I redacted a few personal identifiers from the attached logs before posting publicly,
including my local username, hostname, home directory path, MAC addresses, and GPU
UUID. Driver versions, kernel logs, PCI IDs, GPU model, power states, clocks,
temperatures, and the relevant nvidia-powerd/NVIDIA messages were left intact.

combined-nvidia-powerd-evidence.txt (574.6 KB)

gpu-ac-transition-active.csv.txt (653.8 KB)

gpu-ac-transition-summary.txt (3.3 KB)

nvidia-bug-report.log.gz (743.8 KB)

Also wanted to pop in and confirm doing a sudo systemctl restart nvidia-powerd fixes power draw issues on my Lenovo Legion Gen 7. My flow has been even more simple:

  1. Start laptop in a non-performance mode (Auto or Quiet)
  2. Switch laptop to performance mode
  3. Start game
  4. Watch power draw, but it never reaches above the default power limit
  5. nvidia-smi reports a higher than default max available:
    GPU Ceiling Power Limit
    Current Power Limit : 175.00 W
    Requested Power Limit : N/A
    Default Power Limit : 115.00 W
    Min Power Limit : 1.00 W
    Max Power Limit : 175.00 W

If I perform a sudo systemctl restart nvidia-powerd between steps 2 and 3 (restarting the service AFTER switching laptop to performance mode) OR even restarting the service while a game is running with performance mode active, power draw exceeds 115w.

Edit to add:

NVIDIA-SMI 615.71.09
KMD Version: 615.71.09
CUDA UMD Version: 13.4
xorg-x11-drv-nvidia-power-615.71.09-3.fc44.x86_64

Operating System: Ultramarine Linux 44
KDE Plasma Version: 6.7.5
KDE Frameworks Version: 6.30.0
Qt Version: 6.11.2
Kernel Version: 7.2.5-200.fc44.x86_64 (64-bit)
Graphics Platform: Wayland
Processors: 24 × 12th Gen Intel® Core™ i9-12900HX
Memory: 64 GiB of RAM (62.5 GiB usable)
Graphics Processor 1: NVIDIA GeForce RTX 3080 Ti Laptop GPU
Graphics Processor 2: Intel® UHD Graphics
Manufacturer: LENOVO
Product Name: 82TD
System Version: Legion 7 16IAX7

Running nvidia-smi prints out the current power cap and is reliable to let me know when it has switched over to the higher cap. However, it does require restarting the powerd service to change the power cap. I can not get it to dynamically update when changing my laptop’s power profile in any direction quiet/auto → performance or performance → auto.

I am hit by the same problem (Alienware 18 Area-51 AA18250 with a 5080, Gentoo Linux).

For now, I can reproduce with both drivers 595.99.02 and 615.71.09. Power limits are too low right after boot (80 W instead of 175 W), and sometimes also after profile switching, in some cases profile switch works… This smells like some race condition.

This issue was not present from the start with this setup. I can not fully disentangle whether this started when upgrading to a 7.1.x/7.2.x kernel or after a BIOS update, but the kernel seems more likely.

Same issue here, triggered by booting on battery. Driver 595.104.02 (open), RTX 4070 Laptop.

Setup: ASUS TUF Gaming A15 FA507UI (Ryzen 9 8945H + RTX 4070 Laptop GPU, BIOS FA507UI.310), NixOS, kernel 7.2.8, open kernel modules 595.104.02, nvidia-powerd running (Dynamic Boost enabled).

What I tried:

  1. Switching the ASUS platform profile (performance → balanced → performance) to make the firmware resend its power state: no change, still 900 MHz.
  2. sudo systemctl restart nvidia-powerd while the game was running: fixed it instantly. Clocks jumped to 2310–2370 MHz, the throttle reasons cleared (only “idle” left, because of the game’s 60 fps cap), and utilization dropped to about 40% for the same frame rate.

What I checked on the firmware/driver side:

  • The ACPI tables have only one device with a _PSR method (_SB.ACAD).
  • The EC handler for plug/unplug (_Q3F) runs Notify(ACAD, 0x80), which is the event the open kernel module hooks in nv-acpi.c.
  • The kernel saw the AC change (ACAD went online), and no NVRM errors were logged at that moment, so the driver doesn’t appear to have failed handling the event.
  • The PlatformRequestHandler failed to get target temp / platform power mode from SBIOS warnings appear on every boot, on battery or not, so I don’t think they’re related.

My guess is that nvidia-powerd sets up its battery-mode limits at startup and doesn’t release them when the AC event arrives later. Restarting it re-reads the power source, which is consistent with the other reports here

Workaround for anyone who needs it: a udev rule that restarts nvidia-powerd whenever the charger state changes.
SUBSYSTEM==“power_supply”, ATTR{type}==“Mains”, ACTION==“change”, RUN+=“/usr/bin/systemctl --no-block try-restart nvidia-powerd.service”
Happy to provide an nvidia-bug-report.log.gz if NVIDIA wants one.

Alienware / Dell fixed this issue on my end with their latest 2.9.0 BIOS for the device. It seems also Windows users were affected, so at least in my case, this was a BIOS / UEFI regression.

Strange that it would “sometimes” work with nvidia-powerd, though.

Very likely the same bug, on a different laptop (Lenovo Legion Pro 5 16ARX8, RTX 4060 Laptop, driver 615.78.08), reproduced 3/3. I traced it at the nvidia-powerd ioctl level and with its own verbose log. On the DC->AC transition JpacWorker logs ‘AC event 1’ and goes quiet, but its stop path (JpacWorker::postStop / stop) runs only on process exit. The 0x2080e0af calls with field[3]=0 happen inside that stop window. Details below, logs attached. All times are CET (UTC+1).

System

  • Lenovo Legion Pro 5 16ARX8 (82WM), BIOS LPCN65WW (03/25/2026), BIOS graphics mode: dGPU only (panel on the NVIDIA GPU)
  • AMD Ryzen 7 7745HX, NVIDIA GeForce RTX 4060 Laptop GPU, VBIOS 95.07.18.00.e0
  • Driver 615.78.08, open kernel modules; nvidia-powerd 2.0 (build 1), enabled
  • NVreg_DynamicPowerManagement=3; kernel 7.3.1 (CachyOS); KDE Plasma 6.8.2, Wayland
  • Platform: cTGP 80 W + Dynamic Boost offset 55 W (enforced.power.limit 135-140 W on AC, 55 W on battery)
  • Platform profile: performance

Steps to reproduce

  1. On AC, run a sustained 3D load (100% GPU utilization). GPU is in P0, memory clock 8001 MHz.
  2. Unplug AC for ~20 s while the load keeps running.
  3. Plug AC back in, keep the load running.

Result: 3/3 attempts (22:34, 22:48 and 22:58 on 2026-11-07; in the third, the load started while on battery). Earlier the same day the state also persisted for 4.5 hours
(from an unplug at 17:50 until nvidia-powerd was restarted at 22:23), through later AC plug/unplug cycles.
Not tested here: plug-in with the GPU idle (the linked thread reports that it recovers in that case).

Expected

After AC returns, the GPU goes back to P0 with the AC power policy.

Actual

  • GPU stays in P3: memory clock 7001 MHz instead of 8001.
  • clocks_event_reasons = 0x24 (SW Power Cap + SW Thermal Slowdown), permanently. Both counters grow exactly 1:1 with
    wall-clock time.
  • enforced.power.limit goes back to 140 W.
  • In the three repros the load was thermally bound (86-88 °C in both states), so power draw was similar stuck and healthy
    (~60-68 W). The measurable difference was P3 / 7001 MHz / 0x24 vs P0 / 8001 MHz / 0x20.
  • The 4.5 h episode, at lower temperature, showed the full impact: 100% utilization at 1155 MHz and ~24 W, 57-61 °C
    (HW Slowdown, HW Thermal, Power Brake inactive).

sudo systemctl restart nvidia-powerd fixes it within ~1 s, even under load:
P3 / 7001 MHz / 0x24 → P0 / 8001 MHz / 0x20 (normal thermal target 87 °C).

Mechanism (ioctl trace of nvidia-powerd, repro2-sequence.txt)

All NV_ESC_RM_CONTROL calls from nvidia-powerd were captured with params before and after each call. Command IDs
not in the open headers are given as raw IDs.

Phase What nvidia-powerd sends
AC, healthy only the periodic sampling calls; PERF_GET_POWERSTATE ~0.5/s
DC entry (22:48:00) 0x2080e0d0 params {1, 9} (status 0x56). 0x2080e0af {3/2, 1, 0xf4/0xf8, 2, …}, then every 2 s. 0x20802068 PERF_GET_CURRENT_PSTATE 2/s. 0x2080a0a8 1/s. 0x13b NVPCF_GET_POWER_MODE_INFO every 4 s. PERF_GET_POWERSTATE 10/s
AC return (22:48:21) 0x13b NVPCF_GET_POWER_MODE_INFO x5 (driver re-reads NVPCF func 1, 2, 9 from SBIOS: AC values). After this, 0x2080e0af / 0x20802068 / 0x2080a0a8 stop. No 0x2080e0af with field[3]=0, no 0x2080e0d0 {0, 9}, no 0x216. PERF_GET_POWERSTATE stays at 10/s with the driver answering AC
Stuck (22:48:21 to 22:49:59) nothing but the periodic sampling calls; GPU in P3 / 0x24
nvidia-powerd SIGTERM (22:49:59.95, old PID, between “Stopping” and “Quit successfully”) 0x216 x4, 0x2080e0d0 {0, 9}, 0x2080e0af {3/2, 1, 0xf4/0xf8, 0, …} → next 1 Hz sample (22:50:00.9) shows P0 / 8001 MHz / 0x20
New instance (22:50:00) normal AC init
  • The last 0x2080e0af with field[3]=2 is at 22:48:20.79.
  • There is no 0x2080e0af call at all between AC return (22:48:21) and exit (22:49:59.95).
  • The DC->AC transition path does not send the exit-time release (0x2080e0af with field[3]=0, 0x2080e0d0 {0, 9}).
  • With 1 Hz GPU sampling I cannot separate the exit-time release from the new instance’s init (~50 ms apart) as
    what restores P0. But see repro3 below: the 0x2080e0af calls at exit fall inside JpacWorker’s stop window.
  • 0x2080e0d0 returns NV_ERR_NOT_SUPPORTED (0x56) on this system in every phase. Possibly harmless, but maybe part of
    why the AC path takes a different branch.

nvidia-powerd’s own verbose log (repro3-powerd-verbose.log)

nvidia-powerd only logs to syslog, and its log level is a global set to 2 in main() (non-PIE binary, address 0x6983a8;
76 messages are gated at level >= 3/4). I found no env var or command-line switch that raises it. For the third repro (3/3, 22:57:56 unplug,
22:58:21 plug-in, GPU load started on battery and kept through the plug-in) I set it to 4 with gdb in the running process.

  • Battery: INFO: Switched from AC to battery, JpacWorker AC event 0., JpacWorker NVPCF event 1., then one
    INFO: JPAC prevGpcclkMHz: ..., fpsCap ..., igp.getTgpPwrAvgW(): ..., ljp2.m_limit / 1000.0f: ... line every second.
  • AC return: INFO: Switched from battery to AC, JpacWorker AC event 1., JpacWorker NVPCF event 1.
    • After this, the JPAC lines stop; only JpacWorker::update() once per second.
    • No JpacWorker::postStop() / JpacWorker stop... and no other message about releasing the limit.
    • GPU stays in P3 / 0x24.
  • Exit (systemctl restart, 22:59:32): systemState stop 1, Uninitialize RM Client, JpacWorker::postStop(),
    JpacWorker stop..., JpacWorker stop completed., JpacWorker::postUninitialize(), QBoost is STOPPED now.
    → GPU back to P0 / 8001 MHz.

In the same run the ioctl trace was active:

  • JpacWorker::postStop() at 22:59:32.5394, then “stop completed” at 22:59:32.5473.
  • The two 0x2080e0af calls at exit fall inside that window, at 22:59:32.5397 and 22:59:32.5415.

So on the AC event JPAC stops updating, but its stop path, which issues those 0x2080e0af calls, runs only on process exit.

Frame-rate limiter is released correctly (repro3-frl-packets.txt, D-Bus nvidia.powerd.frl.limit / Frlpacket):

  • On battery, a packet every 2 s with the frame-time limit, for example 16881 us = 59.2 FPS = fpsCap 59.2 in the log.
  • At AC return (22:58:22.324), a 0 0 packet is sent.
  • So the FRL side is released on AC. The RM-side request (0x2080e0af) is what stays behind.

Other evidence

  • PERF_GET_POWERSTATE (0x2080205a) answers (repro-timeline.txt): BATTERY on unplug, AC from ~1 s after plug-in.
    The driver learns AC correctly.
  • ACPI path is complete on both transitions:
    • EC query → Notify(ADP0, 0x80) → nv_acpi_powersource_hotplug_event(0x80) → Notify(NPCF, 0xC0) → nv_acpi_nvpcf_event(0xC0).
    • No GPU D-notifier (Notify PEGP 0xD1-0xD5) on either transition; EC D-notifier level is D1.
  • SBIOS NVPCF replies (nv_acpi_dsm_method, repro2-sequence.txt). Func 2 dynamic params:
    • DC: TGPA 0x0000, TPPA 0x0280, MAGA 0x00c8.
    • AC: TGPA 0x0208, TPPA 0x01b8, MAGA 0x00a0.
    • DROP 0x88b8 in both.
    • Func 9 polled every 4 s on DC.
    • Lenovo WMI: cTGP 80 W, DB offset 55 W. Per the DSDT logic (custom mode off), GPS func 0x2A gives a thermal target of 87 °C. This was not captured at runtime.
  • At its default log level nvidia-powerd logs nothing beyond “nvidia-powerd version:2.0 (build 1)” /
    “DBus Connection is established”, also with __RM_ENABLE_VERBOSE_OUTPUT=1 in its environment.
    /var/log/nvtopps/ exists but stays empty; strace shows the daemon only stat()s it and logs via /dev/log.
  • At driver load, on every boot: “PlatformRequestHandler failed to get target temp from SBIOS” and
    “failed to get platform power mode from SBIOS” (NV_ERR_INVALID_DATA). The GPU works normally before the AC event,
    so possibly unrelated.

Notes

  • nvidia-powerd.service ships with StartLimitIntervalSec=infinity and StartLimitBurst=5.
    • So restart-based workarounds (such as the udev rule suggested in the linked thread) hit start-limit-hit after
      five starts per boot.
    • We override it with StartLimitIntervalSec=60.
  • Workaround in use here: a udev rule on power_supply Mains “change” that starts a oneshot unit. The unit waits 5 s
    (to coalesce adapter flapping) and runs systemctl try-restart nvidia-powerd.

Attachments

All bug reports anonymized:

  • Hostname, user name, serial numbers, UUIDs, MAC/IP addresses and sudo lines replaced.
  • Monitor EDIDs, the plasmashell journal and the encrypted NVML log removed.
  • nvidia-debugdump archive kept as is.

All files are in the attached nvidia-powerd-ac-bug-report.tar.gz (folder nvidia-report/, also contains this text as REPORT.md):

  • nvidia-bug-report-stuck.log.gz: taken while stuck (first repro, 22:35, between plug-in and restart).
  • nvidia-bug-report-healthy.log.gz: taken in the healthy state after the second repro (22:50), for diffing.
  • repro-timeline.txt: first repro. 1 Hz GPU state, ACPI/EC/driver notify trace, PERF_GET_POWERSTATE rate and answers.
  • repro3-powerd-verbose.log: third repro, nvidia-powerd journal at log level 4, from unplug to exit.
  • repro3-frl-packets.txt: third repro, decoded Frlpacket D-Bus signals.
  • repro2-sequence.txt: second repro. Every non-periodic nvidia-powerd RM control with params in/out and status,
    SBIOS _DSM replies, ACPI events, 1 Hz GPU state, including the exit-time release sequence.

nvidia-powerd-ac-bug-report.tar.gz (1.3 MB)

Update: still reproduces on 620.70.06 (kernel 7.3.3, fresh boot, nvidia-powerd still reports 2.0 (build 1)).

  • Load: Blender headless EEVEE render at 4K, 97-100% GPU. Unplug at 23:31:05.7, plug-in at 23:31:30.6 (CET).
  • After AC returns, the GPU stays in P4 with memory clock 6001 MHz (on 615.78.08 it was P3 / 7001), clocks_event_reasons 0x24 / 0x424.
  • Same nvidia-powerd signature as on 615:
    • 0x2080e0af with field[3]=2 every 2 s on battery, the last one at 23:31:29.15, none after AC returns;
    • PERF_GET_POWERSTATE polled at 10/s with the driver answering AC.
  • systemctl restart nvidia-powerd at 23:32:34.318: the exiting process sends 0x2080e0af with field[3]=0 (23:32:34.333 / .336), and the GPU is back at P0 / 8001 MHz / 0x20 on the next sample.

Timeline attached (repro620-timeline.txt).

repro620-timeline.txt (13.0 KB)