### 显存与功耗墙  （2026-10-09T21:29:45+08:00）
### rocm-smi --showmeminfo vram
GPU[0]		: VRAM Total Memory (B): 17163091968
GPU[0]		: VRAM Total Used Memory (B): 206479360

### rocm-smi --showtemp
====================================== Temperature =======================================
GPU[0]		: Temperature (Sensor edge) (C): 34.0
GPU[0]		: Temperature (Sensor junction) (C): 36.0
GPU[0]		: Temperature (Sensor memory) (C): 33.0

### rocm-smi --showpower
=================================== Power Consumption ====================================
GPU[0]		: Current Socket Graphics Package Power (W): 15.0

### rocm-smi --showpowercap  （ROCm 6.2 不支持该选项，原样记录其报错）
usage: rocm-smi [-h] [-V] [-d DEVICE [DEVICE ...]] [--alldevices] [--showhw] [-a] [-i] [-v] [-e [EVENT ...]]
                [--showdriverversion] [--showtempgraph] [--showfwinfo [BLOCK ...]] [--showmclkrange] [--showmemvendor]
                [--showsclkrange] [--showproductname] [--showserial] [--showuniqueid] [--showvoltagerange] [--showbus]

### 功耗墙改从 hwmon 读取（AMD 卡 = 有 power1_cap 的那个 hwmon）
  /sys/class/drm/card2/device/hwmon/hwmon3:
    power1_cap         = 190000000 uW
    power1_cap_default = 190000000 uW
    power1_cap_max     = 190000000 uW
    power1_cap_min     = 0 uW
    power1_label       = PPT
    temp1_label        = edge
