Baseline

Those tests were performed in a full multi-user mode Wayland graphical session, with this command:

systemctl --user stop swayidle syncthing ; sudo rfkill block all ; brightnessctl set 0% ; sudo powerstat ; brightnessctl set 50%

No modules, no radios, no backlight, screen off

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.8   0.0  1.0  403.6  213.8  5.5  3.4  5.8   2.27 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  370.3  203.0  0.0  0.0  0.0   2.26 
  StdDev   0.0   0.1   0.1   0.2   0.0  0.0  284.4  107.2 15.8  9.4 16.4   0.24 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  98.6   0.0  1.0  310.7  175.6  0.0  0.0  0.0   2.11 
 Maximum   0.1   0.7   0.4  99.9   0.1  1.0 1926.3  788.5 81.0 43.0 81.0   3.24 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Seems close to our typical 2.11 baseline, let's keep a 2.1W baseline.

But we can already see we don't have a major power usage improvement here.

1 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.8   0.0  1.0  427.4  233.8  6.0  3.6  6.4   2.16 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  366.8  215.2  0.0  0.0  0.0   2.15 
  StdDev   0.0   0.2   0.1   0.3   0.0  0.2  453.0  162.3 16.9  9.7 15.9   0.25 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  98.3   0.0  1.0  297.2  178.4  0.0  0.0  0.0   2.04 
 Maximum   0.1   1.0   0.5  99.9   0.1  2.0 2858.7 1105.1 82.0 43.0 81.0   3.45 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

This is lower than our baseline, which is certainly an improvement. It's possible the USB-A power usage problem has been fixed, although it was not a big problem when idle (10mW)...

2 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.8   0.0  1.0  428.1  227.8  7.0  4.4  7.2   2.23 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  366.0  207.4  0.0  0.0  0.0   2.22 
  StdDev   0.0   0.2   0.1   0.3   0.0  0.2  466.5  172.6 27.2 17.4 26.3   0.24 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  98.2   0.0  1.0  304.8  174.1  0.0  0.0  0.0   2.07 
 Maximum   0.1   1.0   0.6  99.9   0.1  2.0 2936.3 1155.1 152.0 97.0 148.0   3.38 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Still lower than baseline, starting to question the baseline.

3 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.3   0.0   0.2  99.4   0.0  1.1  494.6  276.4 23.7 15.4 23.6   2.80 
 GeoMean   0.1   0.0   0.1  99.4   0.0  1.1  410.5  233.6  0.0  0.0  0.0   2.64 
  StdDev   0.9   0.1   0.3   1.2   0.0  0.3  447.1  224.4 84.3 49.8 82.7   1.19 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  94.6   0.0  1.0  307.0  174.4  0.0  0.0  0.0   2.14 
 Maximum   4.3   0.8   1.1  99.9   0.1  2.0 2262.6 1023.6 456.0 251.0 447.0   7.20 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

More reasonable.

Ethernet

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.8   0.0  1.0  466.6  240.2  6.1  3.5  6.1   3.70 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  365.6  210.7  0.0  0.0  0.0   3.70 
  StdDev   0.0   0.2   0.1   0.4   0.0  0.0  692.6  230.4 17.1  9.8 15.9   0.23 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  97.9   0.0  1.0  304.3  169.5  0.0  0.0  0.0   3.60 
 Maximum   0.3   1.0   0.7  99.9   0.1  1.0 4189.7 1478.2 83.0 44.0 82.0   4.91 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

This result (+1.5W) Seems to match previous results (+1.6W).

DisplayPort

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.8   0.0  1.1  442.3  235.8  5.8  3.5  6.3   2.58 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  370.5  213.8  0.0  0.0  0.0   2.57 
  StdDev   0.0   0.2   0.1   0.3   0.0  0.2  523.7  184.5 16.3  9.6 15.8   0.30 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  98.2   0.0  1.0  310.0  189.2  0.0  0.0  0.0   2.36 
 Maximum   0.1   1.0   0.6  99.9   0.1  2.0 3256.6 1227.6 82.0 43.0 81.0   3.94 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

That's about +200mW above baseline, similar to our previous results, so not a big improvement in power savings there.

1 HDMI

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.3   0.0   0.2  99.4   0.0  1.1  503.7  286.3 27.1 17.6 27.2   3.07 
 GeoMean   0.1   0.0   0.1  99.4   0.0  1.0  406.0  239.8  0.0  0.0  0.0   2.96 
  StdDev   0.9   0.2   0.3   1.2   0.0  0.2  524.5  243.5 84.5 50.5 82.1   1.01 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  94.6   0.0  1.0  312.9  185.2  0.0  0.0  0.0   2.49 
 Maximum   4.2   1.1   1.2  99.9   0.1  2.0 2833.8 1106.1 450.0 251.0 436.0   6.91 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

About +400mW above baseline, similar to our previous results (+500mW) as well.

2 HDMI

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.8   0.0  1.0  428.2  229.4  9.4  5.6  9.4   3.03 
 GeoMean   0.0   0.0   0.1  99.8   0.0  1.0  374.9  212.0  0.0  0.0  0.0   3.02 
  StdDev   0.0   0.1   0.1   0.2   0.0  0.2  382.5  140.9 23.1 13.5 23.2   0.25 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  98.6   0.0  1.0  308.1  180.7  0.0  0.0  0.0   2.84 
 Maximum   0.1   0.8   0.5  99.9   0.1  2.0 2431.2  948.4 108.0 64.0 108.0   3.92 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Seems to confirm a +400mW, which is similar to the previous results, but actually would be an improvement of 100mW, non-negligible.

2 USB-C

After the above tests, I tried to reproduce the "no power usage" results with 2 USB-C cards plugged in:

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  316.6  186.6  4.2  2.4  4.5   2.73 
 GeoMean   0.0   0.0   0.0  99.9   0.0  1.0  315.5  186.1  0.0  0.0  0.0   2.73 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0   27.6   13.9 13.0  6.8 13.0   0.04 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.7   0.0  1.0  288.1  171.6  0.0  0.0  0.0   2.68 
 Maximum   0.1   0.0   0.1  99.9   0.1  1.0  424.2  247.9 72.0 36.0 72.0   2.85 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Anomalous, trying again, this time with 3 USB-C cards...

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.1  384.1  215.0 19.5 11.7 19.2   2.94 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.1  343.8  197.9  0.0  0.0  0.0   2.91 
  StdDev   0.5   0.0   0.2   0.7   0.0  0.3  306.3  145.8 87.9 50.4 84.6   0.52 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  95.9   0.0  1.0  291.1  170.5  0.0  0.0  0.0   2.69 
 Maximum   2.9   0.2   1.1  99.9   0.1  2.0 1995.4  993.0 491.0 281.0 473.0   4.68 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Oh. Firefox emacs running. Killed.

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  306.1  181.7  8.9  5.6  9.6   2.71 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  302.6  181.0  0.0  0.0  0.0   2.71 
  StdDev   0.0   0.0   0.0   0.1   0.0  0.0   52.5   17.6 23.4 13.9 23.4   0.04 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.6   0.0  1.0  270.5  163.3  0.0  0.0  0.0   2.67 
 Maximum   0.1   0.2   0.1  99.9   0.1  1.0  521.8  258.2 108.0 64.0 108.0   2.88 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Still anomalous. Retrying baseline without any cards, after reboot:

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  330.7  201.2  8.7  4.7  9.7   2.03 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  326.8  200.2  0.0  0.0  0.0   2.03 
  StdDev   0.0   0.0   0.0   0.1   0.0  0.0   58.6   21.2 24.6 14.0 25.3   0.05 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.5   0.0  1.0  301.6  183.2  0.0  0.0  0.0   1.97 
 Maximum   0.1   0.2   0.1  99.9   0.1  1.0  571.7  288.6 117.0 71.0 122.0   2.25 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Conclusion: something was triggered during the tests, invalidating the above results. It turns out I still had various timers firing in my session, and rfkill block doesn't actually do what I would expect...

Redoing baseline

So we start again after the above reboot, with a more complete session cleanup:

systemctl --user stop mbsync.timer notmuch-new.timer swayidle syncthing ; sudo rfkill block all ; brightnessctl set 0% ; sudo powerstat ; brightnessctl set 50%

1 USB-C

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  319.0  194.7  5.3  3.4  5.8   2.01 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  317.8  194.3  0.0  0.0  0.0   2.01 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0   30.4   13.0 19.4 11.6 19.4   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.8   0.0  1.0  298.8  175.4  0.0  0.0  0.0   1.96 
 Maximum   0.1   0.0   0.1  99.9   0.0  1.0  459.2  241.1 108.0 64.0 108.0   2.11 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

That's more like it.

2 USB-C

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.1  376.3  222.4 18.1 10.5 17.8   2.16 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.0  341.5  209.4  0.0  0.0  0.0   2.13 
  StdDev   0.4   0.0   0.1   0.6   0.0  0.2  250.3  108.7 70.5 42.8 66.6   0.43 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  96.8   0.0  1.0  296.4  183.4  0.0  0.0  0.0   1.95 
 Maximum   2.4   0.0   0.7  99.9   0.1  2.0 1497.8  703.0 380.0 237.0 352.0   3.69 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Again, minimum below baseline, which is what we expect.

3 USB-C

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.1  376.3  222.4 18.1 10.5 17.8   2.16 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.0  341.5  209.4  0.0  0.0  0.0   2.13 
  StdDev   0.4   0.0   0.1   0.6   0.0  0.2  250.3  108.7 70.5 42.8 66.6   0.43 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  96.8   0.0  1.0  296.4  183.4  0.0  0.0  0.0   1.95 
 Maximum   2.4   0.0   0.7  99.9   0.1  2.0 1497.8  703.0 380.0 237.0 352.0   3.69 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

So now we are back to our original baseline, and possibly better, even, at 1.95W. It should be noted that the original 2.1W baseline was a bit arbitrary and we were seeing 1.96W power usage results in some tests, so this is not actually a baseline improvement.

New baseline: 1.95W

In any case, we should consider all results apart from this baseline to be invalid, and we will consider the new baseline to be 1.95W.

So let's look at the other expansion cards again.

1TB SSD

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  314.2  193.1  5.3  3.4  5.8   3.85 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  313.3  192.9  0.0  0.0  0.0   3.85 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.2   26.9    8.6 19.4 11.6 19.1   0.20 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.8   0.0  1.0  291.3  182.0  0.0  0.0  0.0   3.72 
 Maximum   0.1   0.0   0.2  99.9   0.1  2.0  449.3  229.0 108.0 64.0 107.0   4.62 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

That is 1.6W above the 2.1W baseline, or 1.8W above the new 1.95W baseline. This is similar to previous results, so no improvement there.

Before resuming tests here, make sure we validate a baseline.

MicroSD

1 USB-A

2 USB-A

3 USB-A

Ethernet

DisplayPort

1 HDMI

2 HDMI

2 usbc 1 usb a ethernet

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  318.6  200.1  7.8  4.5 10.6   3.94 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  317.0  199.7  0.0  0.0  0.0   3.94 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0   34.6   13.1 22.7 12.9 23.8   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.8   0.0  1.0  291.9  182.1  0.0  0.0  0.0   3.89 
 Maximum   0.1   0.0   0.1  99.9   0.1  1.0  447.0  234.4 108.0 64.0 108.0   4.01 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

baseline check

Removed all the modules to make sure we were close to baseline:

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  294.8  170.5  4.0  2.4  4.6   1.88 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  294.0  170.3  0.0  0.0  0.0   1.88 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0   24.4    8.4 13.1  6.8 13.1   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.8   0.0  1.0  276.9  158.9  0.0  0.0  0.0   1.82 
 Maximum   0.1   0.0   0.1  99.9   0.0  1.0  419.7  203.3 72.0 36.0 72.0   1.94 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Turns out we're actually pretty far below the baseline, again. Fascinating.

1 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.0  373.2  208.0 18.0 10.6 17.6   2.18 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.0  335.5  191.2  0.0  0.0  0.0   2.14 
  StdDev   0.5   0.0   0.2   0.7   0.0  0.2  297.3  145.0 86.1 49.0 83.5   0.53 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  95.9   0.0  1.0  299.4  170.9  0.0  0.0  0.0   1.97 
 Maximum   2.9   0.0   1.1  99.9   0.1  2.0 1947.3  984.7 481.0 274.0 467.0   4.02 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Seems to be zero or minimal power usage, depending on the baseline, let's try with more...

2 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  314.1  180.3  5.2  3.3  5.8   2.00 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  313.2  180.0  0.0  0.0  0.0   2.00 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0   26.9    9.8 19.4 11.6 19.1   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.8   0.0  1.0  294.6  164.0  0.0  0.0  0.0   1.97 
 Maximum   0.1   0.0   0.1  99.9   0.1  1.0  451.9  220.2 108.0 64.0 107.0   2.08 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Nice! Exactly the same as before. Let's check with a third one:

3 USB-A

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  305.7  177.6  1.6  1.2  4.0   1.99 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  305.6  177.4  0.0  0.0  0.0   1.99 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0    7.4    9.9  3.4  2.6 11.8   0.02 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.8   0.0  1.0  291.7  166.9  0.0  0.0  0.0   1.94 
 Maximum   0.0   0.0   0.1  99.9   0.1  1.0  331.5  225.5 10.0  7.0 65.0   2.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

So we have confirmation that USB-A expansion cards do not take up more power with the 3.06 BIOS update, I believe.

microsd with sd card

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.0  371.3  205.5 21.4 12.7 21.6   3.72 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.0  335.9  190.2  0.0  0.0  0.0   3.70 
  StdDev   0.4   0.0   0.1   0.6   0.0  0.2  252.8  120.8 63.2 37.5 62.6   0.37 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  97.4   0.0  1.0  288.1  163.0  0.0  0.0  0.0   3.53 
 Maximum   1.8   0.0   0.7  99.9   0.1  2.0 1557.6  793.1 269.0 171.0 262.0   5.23 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 

I had a 128GB micro SD card plugged in during this test.

microsd without card

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.0  322.7  180.2  7.7  4.5  8.1   3.58 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  320.8  179.6  0.0  0.0  0.0   3.58 
  StdDev   0.0   0.0   0.0   0.1   0.0  0.0   39.1   14.5 22.9 12.9 23.5   0.04 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.6   0.0  1.0  302.6  168.6  0.0  0.0  0.0   3.54 
 Maximum   0.1   0.2   0.1  99.9   0.1  1.0  496.9  230.9 109.0 64.0 109.0   3.71 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Seems like the card itself is not changing the minimum, but it is changing the average and maximum significantly.

DisplayPort

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.1  307.5  177.0  1.6  1.2  2.0   2.31 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  307.5  176.9  0.0  0.0  0.0   2.31 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.2    6.6    4.9  3.4  2.6  3.7   0.02 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.8   0.0  1.0  296.3  167.2  0.0  0.0  0.0   2.28 
 Maximum   0.1   0.0   0.1  99.9   0.1  2.0  327.6  187.8  9.0  7.0 11.0   2.37 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

This is similar to our previous results and confirms the DisplayPort card takes a solid 300mW at least. The max and standard deviation were bigger in the previous results, but that could be the result of this test run being more reliable, with the baseline being lower and less flaky.

1 HDMI

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.1   0.0   0.1  99.7   0.0  1.0  370.4  205.7 21.4 12.7 21.4   2.69 
 GeoMean   0.0   0.0   0.1  99.7   0.0  1.0  335.4  190.8  0.0  0.0  0.0   2.66 
  StdDev   0.4   0.0   0.2   0.6   0.0  0.2  251.8  118.7 63.1 36.8 61.9   0.46 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  97.4   0.0  1.0  285.0  160.4  0.0  0.0  0.0   2.43 
 Maximum   1.8   0.0   0.7  99.9   0.1  2.0 1570.0  784.7 261.0 160.0 250.0   4.53 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Again, similar results than before, an extra 400-500mW for a HDMI card, but let's try with two and see what happens...

2 HDMI

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.1  301.9  171.5  7.8  4.6  8.1   2.59 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.1  300.1  171.1  0.0  0.0  0.0   2.59 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.3   36.3   11.9 22.8 13.0 22.6   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.7   0.0  1.0  278.8  158.0  0.0  0.0  0.0   2.53 
 Maximum   0.1   0.0   0.1  99.9   0.1  2.0  443.4  215.6 108.0 64.0 107.0   2.67 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

Now that is an improvement however. We were previously adding more power as HDMI cards were added, but this result seems to show the power usage is constant with more cards.

In other words, the cards are mostly passive, apart from some base circuitry that starts up when any HDMI card is plugged in.

back to baseline: success

Just to double-check, we ran a final baseline test:

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.1  99.9   0.0  1.1  310.4  176.1  4.1  2.4  4.2   2.01 
 GeoMean   0.0   0.0   0.1  99.9   0.0  1.0  309.7  175.9  0.0  0.0  0.0   2.01 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.2   23.2    8.7 13.0  6.8 13.1   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.1  99.7   0.0  1.0  294.6  165.8  0.0  0.0  0.0   1.97 
 Maximum   0.1   0.0   0.1  99.9   0.1  2.0  430.8  215.7 72.0 36.0 72.0   2.12 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

This confirms that we're still close to our baseline and that the results were due to the cards and not external factors.

external USB key

As a comparison point for the MicroSD card reader, here's a power test with an external USB key (Kingston DataTraveler 3.0) plugged in the USB-A port:

-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Average   0.0   0.0   0.0  99.9   0.0  1.0  311.4  178.0  1.6  1.2  4.5   3.89 
 GeoMean   0.0   0.0   0.0  99.9   0.0  1.0  311.3  177.8  0.0  0.0  0.0   3.89 
  StdDev   0.0   0.0   0.0   0.0   0.0  0.0    9.1    8.9  3.4  2.6  9.2   0.03 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------ 
 Minimum   0.0   0.0   0.0  99.9   0.0  1.0  298.8  166.2  0.0  0.0  0.0   3.85 
 Maximum   0.0   0.0   0.1  99.9   0.1  1.0  341.4  211.8 10.0  7.0 48.0   3.94 
-------- ----- ----- ----- ----- ----- ---- ------ ------ ---- ---- ---- ------

So an external USB stick takes up more power than the MicroSD controller, even with a MicroSD card plugged in. It's kind of unfortunate the MicroSD controller doesn't go idle when not in use, because then it would be a great improvement over, say, USB keys.

You can use your Mastodon account to reply to this post.

Created . Edited .