TS-7970 Power Consumption: Difference between revisions
From embeddedTS Manuals
(Created page with "The i.MX6 power consumption can vary a lot depending on the build and activity of the board. Most of the power savings happens automatically when the cpu and gpu are idle, bu...") |
No edit summary |
||
Line 28: | Line 28: | ||
|- | |- | ||
| CPU 100% + GPU loaded (LCD 100%) + IO + Ethernet + HDMI | | CPU 100% + GPU loaded (LCD 100%) + IO + Ethernet + HDMI | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU 100% | | CPU 100% | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU Idle | | CPU Idle + HDMI | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU Idle | | CPU Idle + CPU Ethernet | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU Idle | | CPU Idle | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU Idle | | CPU Idle USB HUB off | ||
| | | | ||
| | |||
|- | |- | ||
| CPU Idle | | CPU Idle USB HUB off, Ethernet PHY in reset | ||
| | | | ||
| | | | ||
|- | |- | ||
| CPU | | Using onboard uC to sleep CPU | ||
| | | | ||
| | | | ||
|} | |} | ||
Line 74: | Line 74: | ||
| (1.23 A) | | (1.23 A) | ||
| (1.08 A) | | (1.08 A) | ||
|- | |- | ||
| CPU Idle + HDMI | | CPU Idle + HDMI | ||
Line 102: | Line 90: | ||
| (0.72 A) | | (0.72 A) | ||
| (0.57 A) | | (0.57 A) | ||
|- | |||
| CPU Idle | |||
| (0.97 A) | |||
| (0.56 A) | |||
|- | |||
| CPU Idle USB HUB off | |||
| (0.70 A) | |||
| (0.55 A) | |||
|- | |||
| CPU Idle USB HUB off, Ethernet PHY in reset | |||
| (0.66 A) | |||
| (0.48 A) | |||
|- | |||
| Using onboard uC to sleep CPU | |||
| (37 mA) | |||
| (21 mA) | |||
|} | |} |
Revision as of 15:35, 30 August 2016
The i.MX6 power consumption can vary a lot depending on the build and activity of the board. Most of the power savings happens automatically when the cpu and gpu are idle, but it is also possible to disable the Ethernet PHY for additional savings.
# Put ETH PHY in reset
echo 116 > /sys/class/gpio/export
echo high > /sys/class/gpio/gpio116/direction
# Put USB HUB in reset
echo 43 > /sys/class/gpio/export
echo low > /sys/class/gpio/gpio43/direction
# Lower backlight to 50%
echo 4 > /sys/class/backlight/backlight_local_lcd/brightness
# Disable backlight
echo 0 > /sys/class/backlight/backlight_local_lcd/brightness
Ethernet is not connected unless otherwise specified, and serial is disconnected during the measurement. The CPU test is 5x processes of "openssl speed", and the GPU test is Qt5CinematicExperience in the Yocto image.
These tests are performed powering the board through 5V.
Test | Max (W) | Average (W) |
---|---|---|
CPU 100% + GPU loaded (LCD 100%) + IO + Ethernet + HDMI | ||
CPU 100% | ||
CPU Idle + HDMI | ||
CPU Idle + CPU Ethernet | ||
CPU Idle | ||
CPU Idle USB HUB off | ||
CPU Idle USB HUB off, Ethernet PHY in reset | ||
Using onboard uC to sleep CPU |
Test | Max (W) | Average (W) |
---|---|---|
CPU 100% + GPU loaded (LCD 100%) + IO + Ethernet + HDMI | (2.16 A) | (1.55 A) |
CPU 100% | (1.23 A) | (1.08 A) |
CPU Idle + HDMI | (0.91 A) | (0.58 A) |
CPU Idle + WIFI on wpa2 running iperf | (1.37 A) | (0.79 a) |
CPU Idle + CPU Ethernet | (1.00 A) | (0.62 A) |
CPU Idle + PCIe Ethernet | (0.72 A) | (0.57 A) |
CPU Idle | (0.97 A) | (0.56 A) |
CPU Idle USB HUB off | (0.70 A) | (0.55 A) |
CPU Idle USB HUB off, Ethernet PHY in reset | (0.66 A) | (0.48 A) |
Using onboard uC to sleep CPU | (37 mA) | (21 mA) |