Tegrastats shows some GPU power consumption even when nothing is running on it(CV does not)

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
1

When I run tegrastats using a 1 min interval, I find that it shows some power to GPU sometimes even though I am not running anything.

CPU always has some power consumption value which is understandable as it is running OS processes. But why would GPU show power at some random time?

For example:
Tegrastat while Running nothing:

GPU utilization zero but GPU is reporting some power consumption.

RAM 3600/15699MB (lfb 923x4MB) CPU [1%@1190,0%@1190,0%@1190,0%@1190,1%@1190,0%@1190,0%@1190,0%@1190] EMC_FREQ 0% GR3D_FREQ 0% AO@39.5C GPU@40C Tboard@40C Tdiode@41.75C AUX@40C CPU@40.5C thermal@40.3C PMIC@100C -----GPU 155/8------ CPU 465/465 SOC 931/784 CV 0/0 VDDRQ 0/0 SYS5V 1574/1421

What is the reason for that, how can I address this random overhead while trying to measure energy consumption caused by an application that I am running?

The DLA and Vision processor ne the other hand, do not show any such behavior.

Thanks.

2

mmonil,

Do you connect monitor during this test?

3

No, I do not connect monitor during the test. I just ssh to the Xavier.

4

Could you disable xorg/gdm during your observation and see if gpu still has power consumption?

5

Hi WayneWWW,
I disabled gdm and made sure no gdm processes are running. Still, I am finding some power consumed by GPU. Currently, I set the Nvpmodel to MaxN, now it is showing 465mW continuously on GPU. Even in default Nvpmmodel(15W), Gpu shows some power consumption. I am trying to measure power consumption by GPU for different application and this default/random value is contributing to the reading as a noise.

However, I am curious to know why this would happen and how can I stop it.

Thanks

6

mmonil,

I will check it on my side.

7

Hi,

Per checked on our device. It looks like gdm is the process that causing background gpu power consumption.
If you want to measure the gpu power, please disable it first.

You could disable it by using

sudo systemctl stop gdm

8

“sudo systemctl stop gdm” stops the gdm but still I am finding some power consumption.
Currently, I am using MAXN for NVPmodel and it is just continuously showing about 465 mwatt.

RAM 2557/15699MB (lfb 320x4MB) CPU [0%@1190,0%@1190,0%@1190,0%@1190,0%@1190,0%@1190,0%@1190,0%@1190] EMC_FREQ 0% GR3D_FREQ 0% AO@26.5C GPU@27.5C Tboard@28C Tdiode@29.5C AUX@27C CPU@27.5C thermal@27.15C PMIC@100C -----GPU 465/465----- CPU 465/465 SOC 776/776 CV 0/0 VDDRQ 0/0 SYS5V 1696/1699

Do you see any process that might be responsible?

11457 XXX 20 0 17548 3976 3088 R 1.0 0.0 0:00.38 top
5303 root 20 0 73824 19156 14216 S 0.3 0.1 2:20.09 sssd_nss
5820 root 20 0 19176 3988 3056 S 0.3 0.0 0:16.31 systemd-logind
1 root 20 0 161596 6780 3784 S 0.0 0.0 1:54.60 systemd
2 root 20 0 0 0 0 S 0.0 0.0 0:07.94 kthreadd
3 root 20 0 0 0 0 S 0.0 0.0 0:21.87 ksoftirqd/0
5 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/0:0H
7 root 20 0 0 0 0 S 0.0 0.0 5:35.16 rcu_preempt
8 root 20 0 0 0 0 S 0.0 0.0 1:26.40 rcu_sched
9 root 20 0 0 0 0 S 0.0 0.0 0:00.08 rcu_bh
10 root rt 0 0 0 0 S 0.0 0.0 0:04.08 migration/0
11 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 lru-add-drain
12 root rt 0 0 0 0 S 0.0 0.0 0:02.71 watchdog/0
13 root 20 0 0 0 0 S 0.0 0.0 0:00.00 cpuhp/0
14 root 20 0 0 0 0 S 0.0 0.0 0:00.00 cpuhp/1
15 root rt 0 0 0 0 S 0.0 0.0 0:03.20 watchdog/1
16 root rt 0 0 0 0 S 0.0 0.0 0:03.23 migration/1
17 root 20 0 0 0 0 S 0.0 0.0 0:04.62 ksoftirqd/1
19 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/1:0H
20 root 20 0 0 0 0 S 0.0 0.0 0:00.00 cpuhp/2
21 root rt 0 0 0 0 S 0.0 0.0 0:03.43 watchdog/2
22 root rt 0 0 0 0 S 0.0 0.0 0:03.92 migration/2
23 root 20 0 0 0 0 S 0.0 0.0 0:09.60 ksoftirqd/2
25 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/2:0H
26 root 20 0 0 0 0 S 0.0 0.0 0:00.00 cpuhp/3
27 root rt 0 0 0 0 S 0.0 0.0 0:02.91 watchdog/3
28 root rt 0 0 0 0 S 0.0 0.0 0:02.89 migration/3
29 root 20 0 0 0 0 S 0.0 0.0 0:07.75 ksoftirqd/3
31 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/3:0H
32 root 20 0 0 0 0 S 0.0 0.0 0:00.08 cpuhp/4
33 root rt 0 0 0 0 S 0.0 0.0 0:02.90 watchdog/4
34 root rt 0 0 0 0 S 0.0 0.0 0:02.76 migration/4
35 root 20 0 0 0 0 S 0.0 0.0 0:03.41 ksoftirqd/4
37 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/4:0H
38 root 20 0 0 0 0 S 0.0 0.0 0:00.05 cpuhp/5
39 root rt 0 0 0 0 S 0.0 0.0 0:02.69 watchdog/5
40 root rt 0 0 0 0 S 0.0 0.0 0:02.45 migration/5
41 root 20 0 0 0 0 S 0.0 0.0 0:03.16 ksoftirqd/5
43 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/5:0H
44 root 20 0 0 0 0 S 0.0 0.0 0:00.09 cpuhp/6
45 root rt 0 0 0 0 S 0.0 0.0 0:02.06 watchdog/6
46 root rt 0 0 0 0 S 0.0 0.0 0:02.67 migration/6
47 root 20 0 0 0 0 S 0.0 0.0 0:03.35 ksoftirqd/6
49 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/6:0H
50 root 20 0 0 0 0 S 0.0 0.0 0:00.10 cpuhp/7
51 root rt 0 0 0 0 S 0.0 0.0 0:02.77 watchdog/7
52 root rt 0 0 0 0 S 0.0 0.0 0:02.34 migration/7
53 root 20 0 0 0 0 S 0.0 0.0 0:03.15 ksoftirqd/7
55 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/7:0H
56 root 20 0 0 0 0 S 0.0 0.0 0:00.02 kdevtmpfs
57 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 netns
230 root -51 0 0 0 0 S 0.0 0.0 0:00.00 irq/100-mc_stat
701 root 20 0 0 0 0 S 0.0 0.0 0:03.80 khungtaskd
702 root 20 0 0 0 0 S 0.0 0.0 0:02.77 oom_reaper
703 root 0 -20 0 0 0 S 0.0 0.0 0:02.03 writeback
704 root 20 0 0 0 0 S 0.0 0.0 0:00.00 kcompactd0
706 root 25 5 0 0 0 S 0.0 0.0 0:00.00 ksmd

9

mmonil,

It is common that there are some low power maintain on the rail.

10

Hi WayneWWW,
Thanks for your reply.