Reduce power consumption with powertop


Recommended Posts

I'm also experiencing issues getting my Pkg C-states lower than C2.

I replaced my Mellanox Connect-X3 card to an Intel X710-DA4.

And also changed my LSI 3008 to an SilverStone Technology ECS06 with an ASMedia-ASM1166 chip.

Now my whole system supports ASPM.

 

I'm using a 13900KS with the Asus Z790 Pro Art Mainboard.

I disabled the 10GB internal NIC (no ASPM support unfortunately) and also enabled all C-state options.

 

Output of "lspci -vv | awk '/ASPM/{print $0}' RS= | grep --color -P '(^[a-z0-9:.]+|ASPM )'":

 

Spoiler

00:01.0 PCI bridge: Intel Corporation Device a70d (rev 01) (prog-if 00 [Normal decode])
                LnkCap: Port #2, Speed 32GT/s, Width x8, ASPM L0s L1, Exit Latency L0s unlimited, L1 <16us
                LnkCtl: ASPM L0s L1 Enabled; RCB 64 bytes, Disabled- CommClk-
00:01.1 PCI bridge: Intel Corporation Device a72d (rev 01) (prog-if 00 [Normal decode])
                LnkCap: Port #4, Speed 32GT/s, Width x8, ASPM L0s L1, Exit Latency L0s <4us, L1 <16us
                LnkCtl: ASPM L0s L1 Enabled; RCB 64 bytes, Disabled- CommClk+
00:06.0 PCI bridge: Intel Corporation Device a74d (rev 01) (prog-if 00 [Normal decode])
                LnkCap: Port #5, Speed 16GT/s, Width x4, ASPM L0s L1, Exit Latency L0s <4us, L1 <16us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
00:1a.0 PCI bridge: Intel Corporation Device 7a48 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #25, Speed 16GT/s, Width x4, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
                LnkCtl: ASPM L0s L1 Enabled; RCB 64 bytes, Disabled- CommClk-
00:1b.0 PCI bridge: Intel Corporation Device 7a40 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #17, Speed 8GT/s, Width x4, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
                LnkCtl: ASPM L0s L1 Enabled; RCB 64 bytes, Disabled- CommClk-
00:1b.4 PCI bridge: Intel Corporation Device 7a44 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #21, Speed 16GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
00:1c.0 PCI bridge: Intel Corporation Device 7a38 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #1, Speed 8GT/s, Width x2, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
00:1c.2 PCI bridge: Intel Corporation Device 7a3a (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #3, Speed 8GT/s, Width x1, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled+ CommClk+
00:1c.3 PCI bridge: Intel Corporation Device 7a3b (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #4, Speed 8GT/s, Width x1, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
00:1d.0 PCI bridge: Intel Corporation Device 7a30 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #9, Speed 16GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
02:00.0 SATA controller: ASMedia Technology Inc. ASM1166 Serial ATA Controller (rev 02) (prog-if 01 [AHCI 1.0])
                LnkCap: Port #0, Speed 8GT/s, Width x2, ASPM L0s L1, Exit Latency L0s <4us, L1 <64us
                LnkCtl: ASPM L0s L1 Enabled; RCB 64 bytes, Disabled- CommClk+
03:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd NVMe SSD Controller SM981/PM981/PM983 (prog-if 02 [NVM Express])
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
06:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd NVMe SSD Controller SM981/PM981/PM983 (prog-if 02 [NVM Express])
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
09:00.0 Ethernet controller: Intel Corporation Ethernet Controller I226-V (rev 06)
                LnkCap: Port #0, Speed 5GT/s, Width x1, ASPM L1, Exit Latency L1 <4us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
0a:00.0 Ethernet controller: Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev 01)
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <16us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
0a:00.1 Ethernet controller: Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev 01)
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <16us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
0a:00.2 Ethernet controller: Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev 01)
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <16us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+
0a:00.3 Ethernet controller: Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev 01)
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <16us
                LnkCtl: ASPM L1 Enabled; RCB 64 bytes, Disabled- CommClk+

 

Edited by RedReddington
Link to comment
On 7/29/2023 at 9:43 PM, RedReddington said:

I replaced my Mellanox Connect-X3 card to an Intel X710-DA4.

And also changed my LSI 3008 to an SilverStone Technology ECS06 with an ASMedia-ASM1166 chip.

The only ideas I have left:

- change the whole PCIe settings in the BIOS to PCIe 3.0 (if such a setting exists)

- install the NVMe only on chipset M.2 slots. Does it help?

+ Which NVMe brand / models are you using? Samsung has the best C-State compatibility as of my experience.

- install the Intel 10G in the bottom slot (chipset) and remove the SATA card to test the impact. Does it help? If yes, you could try to buy  an ASM1166 M.2 card and use one of the M.2 chipset slots for this card, too. They idea is not to use CPU PCIe Lanes at all. PS don't use M.2_4 or you loose 4 SATA ports.

- did you disable Thunderbolt in the BIOS (if possible)?

- did you disable audio in the BIOS?

- did you remove the Wifi card?

 

 

 

 

Link to comment

- change the whole PCIe settings in the BIOS to PCIe 3.0 (if such a setting exists)

I can check that, but I think this is only possible for the PCIe devices and not my M.2 nvmes.

 

- install the NVMe only on chipset M.2 slots. Does it help?

I will also check that.

 

- Which NVMe brand / models are you using? Samsung has the best C-State compatibility as of my experience.

I changed both of them to 970 EVO Plus 2TB

 

- They idea is not to use CPU PCIe lanes at all

Using CPU lanes prevents C-states? Is this true?

But I will also remove both my NIC and the ASMedia card and also unplug all sata and nvme devices and boot into a "fresh Unraid" on another USB stick to check the behaviour. Maybe it's on of the usb ports.. I will also try some of them.

 

- did you disable Thunderbolt in the BIOS (if possible)?

- did you disable audio in the BIOS?

- did you remove the Wifi card?

I disabled both audio and thunderbolt. I didn't remove the wifi card, but if i remember correctly I also disabled it.

 

I can check all of this today in the evening, I will report my progress...

Edited by RedReddington
Link to comment

I'm also having issues reducing the actual power consumption of my box. Right now isn't too bad, but living in Germany I want to squeeze every single Watt from the box to a minimum. I've been reading the 18 pages from start to end, and here is my info:

 

Intel i5-13500 13th Gen

ASRock B760m Pro RS/D4

Corsair Vengeance LPX 1x16GB DDR4 3200MHz - Originally planned 4x16, but for testing I have only one.

Noctua NH-L12S

WD SN850x 2TB (also tried with a Samsung I had around with similar result)

Corsair RM850x

 

I've disabled everything unnecessary on the BIOS (LEDs, Audio, Thunderbolt...), activated all the C-States, enabled ASPM for all devices, followed all the recommendations of @mgutt, but cannot get below 30W. Also updated powertop to 2.15 to get the other C-States to show up, but cannot get past C3. Also moved the NVMe from the CPU slot to the Chipset one, no effect. 

 

At this point, I can still return some of the equipment and get another ones, any recommendation on what to do? Seems like the Motherboard is just taking too much power, any recommendations for a mATX board that is actually power efficient?

 

Thanks a lot!

 

 

image.thumb.png.c529168850de5ceb3bc445a1d21d9049.png

Link to comment
9 hours ago, jcgonzalezdelgado said:

Seems like the Motherboard is just taking too much power, any recommendations for a mATX board that is actually power efficient?

A few days ago I tested a Gigabyte Board with under 10W:

https://forums.unraid.net/topic/138498-gigabyte-b760m-gaming-x-ddr4-messung-und-sata-erweiterung/

 

 

And some users tested Kontron boards with under 5W. Check this thread:

https://www.hardwareluxx.de/community/threads/die-sparsamsten-systeme-30w-idle.1007101/

 

 

  • Thanks 1
Link to comment
On 8/1/2023 at 8:11 AM, mgutt said:

A few days ago I tested a Gigabyte Board with under 10W:

https://forums.unraid.net/topic/138498-gigabyte-b760m-gaming-x-ddr4-messung-und-sata-erweiterung/

 

 

And some users tested Kontron boards with under 5W. Check this thread:

https://www.hardwareluxx.de/community/threads/die-sparsamsten-systeme-30w-idle.1007101/

 

 


Thanks a lot! After swapping to the Gigabyte I can reach C8 and my power consumption is half (15-16W)

  • Like 1
Link to comment

Hi, I have installed powertop and have it auto-tune at every boot via go file, all my tunables report "good" but In not able to get passed C3, and even that is only possible with my array stopped, I have a 4790k and with array stopped currently its sat at 120w 

 

Any help would be much appreciated, my electricity bill is killer being in the UK

powertop.png

2023-08-03 11_29_18-Tower_Dashboard.png

tower-diagnostics-20230803-1135.zip

Edited by Fantomm3
Link to comment

I had some problems with my array, but tomorrow I'm up and running again.

So back to my powertop.
With the code of @dopeytree I had some errors, but didn't work at all and didn't give me my webui anymore.

So I tried running powertop on my own, but then still errors, even if I just enter powertop, so how can I resolve them first for getting back at the code of dopeytree?

 

This is the lines I get when running only the "powertop" command:

 

powertop: /lib64/libncursesw.so.6: no version information available (required by powertop)
powertop: /lib64/libtinfo.so.6: no version information available (required by powertop)
modprobe cpufreq_stats failed
Cannot load from file /var/cache/powertop/saved_results.powertop
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only 
RAPL device for cpu 0
RAPL Using PowerCap Sysfs : Domain Mask d
RAPL device for cpu 0
RAPL Using PowerCap Sysfs : Domain Mask d
Devfreq not enabled
glob returned GLOB_ABORTED
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only 

 

Thanks in advance!

Link to comment
On 8/3/2023 at 12:30 PM, Fantomm3 said:

Hi, I have installed powertop and have it auto-tune at every boot via go file, all my tunables report "good" but In not able to get passed C3, and even that is only possible with my array stopped, I have a 4790k and with array stopped currently its sat at 120w 

 

Any help would be much appreciated, my electricity bill is killer being in the UK

powertop.png

2023-08-03 11_29_18-Tower_Dashboard.png

tower-diagnostics-20230803-1135.zip 147.42 kB · 0 downloads

Ive had to enable something extra on my asus z97 board to get it to go to c6, which is the lowest you can achive on the package. 
Aggressive Link Power Management in the Sata settings and various pcie power savings L0sL1s or something, would have too look where exactly that was.

This is not guarantee that it will work afterwards. Compontents could be a factor aswell. Like cheap ssds preventing the whole system from going to c-states at all or a pcie card that doesnt support any power saving features.

Got the 4690k system to sub 20 watts idle tho.

(forgot to mention, same applies to USB devices. I got a rouge Razer mouse which misbehaves for example. USB SSDs/HHDs same as internal, if they dont support powersavings, rip.

Having a monitor plugged in does skew some things too. I got a system where i can leave a VGA plugged in and the system just idles fine but then theres a system that throws a 5 min tantrum until it finally powers down because it finally realized theres no actual image displayed. Digital outputs should work better but who knows)

Edited by Mainfrezzer
Link to comment

When it try the auto command it gives an issue but now when I reload powertop it shows GOOD for everything.  What am I doing wrong?  My cpu is 11600k and gpu is rtx 4060.  Thank you for any help. Right now the rig is using more power than it did even a week ago and I don't know if it's because of nvidia drivers or what.

 

 

powertop 1.png

powertop2.png

powertop3.png

powertop4.png

Link to comment
3 hours ago, rikdegraaff said:

So I know I am not able to use that command, but what setting does cause the parity disk to stop working?

Read the first post and/or search for "ASM". Some controllers have problems and yes you can execute relevant commands individually.

Link to comment
On 5/1/2023 at 11:29 PM, MagickPistacho said:

 

Well, I've changed my board now to a Gigabyte B660M DS3H DDR4 and the RAM to G.Skill Aegis F4 16GB 2400, and it reaches C10, even with the HDMI connected.

 

0008.thumb.png.c4a85396b204a8197468fe645771ae48.png

Right now it sits at 9-10W on spindown, no HDMI, no keyboard or mouse, WebGUI closed. One disc spinning bumps it up to 13-14W, two around 25-35 it seems.

 

I've tried with no discs and it goes down to about 7-8W, and without SSD and fans it makes little to no difference, 6-7 but sometimes 8 as well.

 

It's not bad but this user here managed to reach 4.5W at idle and 9W with 2 HDDs spinning and the same board, CPU, PSU and RAM: https://www.hardwareluxx.de/community/threads/sparsamer-nas-mediaserver-neuling.1329924/#post-29668608

So I don't know what I'm doing wrong, maybe I need to change some settings on BIOS but I haven't found anything that could do the trick.


I have added the build to my list of things to do because its very interesting. Thanks.

 

 

However, I have two questions:
Can you try to install a NVME ?

There are 2 NVME connections, CPU and SB. I would be interested to know what happens if you put an NVME there, because I guess you won't get higher than C8 after using one or both of this slots.

 

 

I would also like to know what settings you changed in the BIOS to lower the consumption.

 

 

Thanks for your effort.

Link to comment
On 8/3/2023 at 12:30 PM, Fantomm3 said:

Hi, I have installed powertop and have it auto-tune at every boot via go file, all my tunables report "good" but In not able to get passed C3, and even that is only possible with my array stopped, I have a 4790k and with array stopped currently its sat at 120w 

 

Any help would be much appreciated, my electricity bill is killer being in the UK

 

 

tower-diagnostics-20230803-1135.zip 147.42 kB · 1 download

No way your server alone is drawing 120W idle with the array powered down. I have a system with a 7700k and it draws 80-100W max. while doing a parity check. There is nothing else connected to the power meter? How do you measure the power draw?

Link to comment
On 8/14/2023 at 10:42 PM, T-Birth said:


I have added the build to my list of things to do because its very interesting. Thanks.

 

 

However, I have two questions:
Can you try to install a NVME ?

There are 2 NVME connections, CPU and SB. I would be interested to know what happens if you put an NVME there, because I guess you won't get higher than C8 after using one or both of this slots.

 

 

I would also like to know what settings you changed in the BIOS to lower the consumption.

 

 

Thanks for your effort.

 

Hey! 🙂

 

Yeah, I've already tested that. In fact right now I have two NVMe (Samsung 970 Evo Plus 2TB) connected to the board and consumption with Powertop is about 9-10W still (3 HDDs, 2 NVMes, 1 USB, 3 fans), around 11-12W without it.


With Powertop it certainly reaches C10.

 

As for settings, it's been a while so I don't really recall and I can't look at it right now, sorry, but basically followed whatever seemed to be usable for other users and did some trial and error to see which of them affected consumption in a positive way.

 

Hope this helps! 

I also bought another Gigabyte board and did more testing with both of the, some RAM modules and options, but had several problems with the new one, this works better. My idea was to post the results here, but these past months have been very rough, I'll see if I can manage to post in the near future.

 

 

  • Thanks 1
Link to comment
On 8/13/2023 at 8:58 PM, mgutt said:

Read the first post and/or search for "ASM". Some controllers have problems and yes you can execute relevant commands individually.

 So which code do i need in my go file?

 

Because there is so much code, I cannot make it fitting for my server.

I don't know which code to use and which not to use.
And scared for parity checks as they take a long time with 18TB...

 

So can someone or @mgutt make me a code that is safe for disabling disks, so that when I put the code in, I dont have to do anything again.
Is that a possibility?

 

My server consists of the following parts:

10 x 3.5" hdds

1 x 2.5" ssd cache

Intel core i3-10100

80GB DDR4 RAM

Gigabyte B560M Aorus ELITE Mobo

Supermicro AOC-SAS2LP-MV8 SATA expansion card

 

Thanks in advance!

 

Link to comment
  • 2 weeks later...
On 8/11/2023 at 6:08 AM, FlyingTexan said:

When it try the auto command it gives an issue but now when I reload powertop it shows GOOD for everything.  What am I doing wrong?  My cpu is 11600k and gpu is rtx 4060.  Thank you for any help. Right now the rig is using more power than it did even a week ago and I don't know if it's because of nvidia drivers or what.

 

 

powertop 1.png

powertop2.png

powertop3.png

powertop4.png

i'm having the same issue unfortunately, and I'm not technical to figure it out sadly. the only thing i can find that's wrong is this when running powertop --help:

 

root@fcloud:~# powertop --help
powertop: /lib64/libncursesw.so.6: no version information available (required by powertop)
powertop: /lib64/libtinfo.so.6: no version information available (required by powertop)
Usage: powertop [OPTIONS]

     --auto-tune         sets all tunable options to their GOOD setting
 -c, --calibrate         runs powertop in calibration mode
 -C, --csv[=filename]    generate a csv report
     --debug             run in "debug" mode
     --extech[=devnode]  uses an Extech Power Analyzer for measurements
 -r, --html[=filename]   generate a html report
 -i, --iteration[=iterations] number of times to run each test
 -q, --quiet             suppress stderr output
 -s, --sample[=seconds]  interval for power consumption measurement
 -t, --time[=seconds]    generate a report for 'x' seconds
 -w, --workload[=workload] file to execute for workload
 -V, --version           print version information
 -h, --help              print this help menu

For more help please refer to the 'man 8 powertop'

root@fcloud:~# apt-get install libncursesw6
bash: apt-get: command not found
root@fcloud:~# 
 

it seems something required by powertop (/lib64/libncursesw.so.6) isnt installed or available on my server. anyone have any idea how to remedy that?

Edited by MichaelVB
additional info
Link to comment
11 hours ago, FlyingTexan said:

And to what I posted? In your first post you say BAD when things are zeroes. I posted my screenshot showing I have all Zeroes. What can I do?

I spent another few hours doing stuff, some maybe helped, some probably didnt, i can;t tell what did or didnt for sure. but i managed to get down to C7, here's what i did (with lots of reboots everytime i did something to make sure it was fresh)

 

- reset all the values in my bios to stock, then carefully went through everything one by one following the original post as much as possible

- delete the "extra" folder from my flash drive to remove powertop

- install NerdTools from the apps store

- installed powertop through NerdTools, this got rid of the "no version information available" errors

- disabled array auto start while i was diagnosing all this stuff to keep all the dockers and background stuff off

- from a fresh boot i ran "powertop --calibrate" (takes a minute or two to test everything it needs to and the puts you back into powertop when it's done and you can hit ESC to see what it did)

- restarted the server

- ran "powertop --calibrate" again, then ran "powertop --auto-tune"

 

now i was getting down to C3, so i started removing hardware to see what the limit was. i removed my M.2 ssd and my GPU and found that the system would go to C7 without the GPU. the M.2 was working with C7. in the end the differene was 6-7 watts so I kept the GPU in.

 

I restarted the array and started dockers one by one, I found my minecraft server would keep the system with 0% in any c state, but all the other dockers allowed it to go down to the max C state my stsem was capable of (either C3 with gpu or C7 without)

 

then i edited the file "go" under the config directory on my flash drive (i have Dynamix file manager installed so used it's text editor in the unraid gui) to include all the bits the mgutt recomended in his guide, except for disabling turbo because i didnt see any differences in idle power consumption on my i3 9100 and i guess why not keep the boost for when it's working at something. so now it would apply all the equivalent to auto tune and setting all the controllers to efficient states at every boot. And voila my 8 hour journey to reduce my server's consumption from 29 to 23 watts was complete lol (got it as low as 15 without case fans or a gpu which is nice. might look at using the IGPU in the future for tdarr and jellyfin transcoding beause each 1W of power is about 2 euros a year in electricity where i live and it could be a fun project)

 

good luck man! and thanks to mgutt for his excellent guide and help to the community ❤️

 

 

Link to comment
3 hours ago, MichaelVB said:

- from a fresh boot i ran "powertop --calibrate" (takes a minute or two to test everything it needs to and the puts you back into powertop when it's done and you can hit ESC to see what it did)

- restarted the server

Everytime you reboot, unRAID is reinstalled. unRAID is not persistent. It's running 100% in the RAM.

 

If you for example install packages through the nerdtools, they are downloaded to the flash drive and freshly installed on every reboot.

  • Like 1
Link to comment
22 hours ago, MichaelVB said:

I spent another few hours doing stuff, some maybe helped, some probably didnt, i can;t tell what did or didnt for sure. but i managed to get down to C7, here's what i did (with lots of reboots everytime i did something to make sure it was fresh)

 

- reset all the values in my bios to stock, then carefully went through everything one by one following the original post as much as possible

- delete the "extra" folder from my flash drive to remove powertop

- install NerdTools from the apps store

- installed powertop through NerdTools, this got rid of the "no version information available" errors

- disabled array auto start while i was diagnosing all this stuff to keep all the dockers and background stuff off

- from a fresh boot i ran "powertop --calibrate" (takes a minute or two to test everything it needs to and the puts you back into powertop when it's done and you can hit ESC to see what it did)

- restarted the server

- ran "powertop --calibrate" again, then ran "powertop --auto-tune"

 

now i was getting down to C3, so i started removing hardware to see what the limit was. i removed my M.2 ssd and my GPU and found that the system would go to C7 without the GPU. the M.2 was working with C7. in the end the differene was 6-7 watts so I kept the GPU in.

 

I restarted the array and started dockers one by one, I found my minecraft server would keep the system with 0% in any c state, but all the other dockers allowed it to go down to the max C state my stsem was capable of (either C3 with gpu or C7 without)

 

then i edited the file "go" under the config directory on my flash drive (i have Dynamix file manager installed so used it's text editor in the unraid gui) to include all the bits the mgutt recomended in his guide, except for disabling turbo because i didnt see any differences in idle power consumption on my i3 9100 and i guess why not keep the boost for when it's working at something. so now it would apply all the equivalent to auto tune and setting all the controllers to efficient states at every boot. And voila my 8 hour journey to reduce my server's consumption from 29 to 23 watts was complete lol (got it as low as 15 without case fans or a gpu which is nice. might look at using the IGPU in the future for tdarr and jellyfin transcoding beause each 1W of power is about 2 euros a year in electricity where i live and it could be a fun project)

 

good luck man! and thanks to mgutt for his excellent guide and help to the community ❤️

 

 

can you tell me how you installed powertop in nerdtools? I'm not seeing it. I don't want to simply start flipping switches in there. 

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.