Read-only hardware snapshot — 18:24:39 Vienna

The ongoing game/test was left untouched. This was one telemetry sample, not a stress test or a hardware-stability diagnosis.

• Windows reports AC power connected, battery at 80%, battery saver off.
• Exact-name lookup found no running MSI Afterburner, RTSS or the listed ASUS GPU Tweak process names. This does not exclude previously applied settings or firmware/OEM tuning.
• Official NVIDIA NvAPI_GPU_GetPstates20 returned success. Graphics and memory frequency offsets are 0 kHz in all exposed P0, P3, P4, P5 and P8 states. No base-voltage or overvoltage entries were exposed. Undervolting and per-point voltage/frequency curves therefore remain unknown.
• nvidia-smi reports P0, GPU 64°C, graphics 2670 MHz, memory 8001 MHz, instantaneous power 34.05 W. Thermal slowdown, hardware power braking and software power capping were not active at this instant.
• Current GPU power ceiling: 111.17 W; default ceiling: 80 W; supported maximum: 130 W. This difference alone does not establish a manual power modification; no setting was changed or inferred to be the cause.
• The driver reports the clock-event reason Reliability: Active. That label alone is not evidence of a TDR or proof of its cause. Default application clocks are reported as deprecated/unavailable.

Evidence: hardware-state.json, nvidia-smi-hardware-state.txt, targeted-tuning-processes.json, and reproducible read_hardware_state.py. The script uses the previously downloaded official NVIDIA NVAPI structure definitions and entry-point IDs. Only read APIs are called.

These readings show no exposed fixed graphics/memory overclock offset and no thermal throttling at the sampled instant. They neither establish complete stock tuning nor distinguish conclusively between application, driver and hardware causes of the earlier resets.