deviceQuery shows GPU Max Clock rate: 72 MHz (0.07 GHz)

Autonomous Machines Jetson & Embedded Systems Jetson TX1
1

After I max out the GPU clock rate, deviceQuery shows very low clock rates:

./deviceQuery/deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GM20B"
  CUDA Driver Version / Runtime Version          7.0 / 7.0
  CUDA Capability Major/Minor version number:    5.3
  Total amount of global memory:                 3854 MBytes (4040933376 bytes)
  ( 2) Multiprocessors, (128) CUDA Cores/MP:     256 CUDA Cores
  GPU Max Clock rate:                            72 MHz (0.07 GHz)
  Memory Clock rate:                             13 Mhz

/sys/kernel/debug/clock/override.gbus/rate:998400000
/sys/kernel/debug/clock/override.gbus/max:998400000
/sys/kernel/debug/clock/override.emc/rate:1600000000
/sys/kernel/debug/clock/override.emc/max:1600000000
/sys/kernel/debug/clock/gbus/rate:998400000
/sys/kernel/debug/clock/gbus/max:998400000
/sys/kernel/debug/clock/emc/rate:1600000000
/sys/kernel/debug/clock/emc/max:1600000000
2

It’s a known issue where deviceQuery is under-reporting lower scaling frequency, but as you confirmed through sysfs, the governor limit has been set to maximum of 998MHz — GPU should run at full speed when engaged under load. If you run nbody sample for example, should be pretty clear that it is working correctly.