Quick update, I switched out my MSI Z690 Carbon for a MSI B760 Pro instead. I had to manually turn on ASPM on the realtek ethernet but I am now down to C8 and 22w idle. Not has good as my old 9th gen HP desktop which idled at about 9w but I can probably live with it. I am still yet to add drives etc and a graphics card for pass through to a VM will likely add 10-15w even at idle unfortunately but worth the trade off of the 12700K CPU upgrade...

Warning, avoid the Kingston KC3000 due to high power usage and high temps in idle!

Setup:

Intel core 12600k (lowered max TDP to 65W due to MB not supporting higher)

Jonsbo N3

2x Arctic P8 Slim PWM PST

2x Noctua NF-A9 PWM

Piesia H670 Motherboard (OEM for example CWWK) ( https://www.alibaba.com/product-detail/Piesia-NAS-Storage-Mainboard-8-SATA_1601298509269.html )

64GB DDR5 4800mhz RAM from Crucial

2x Kingston KC3000 2TB

Corsair SF750 (2024)

This drew 26-27W in idle just after a fresh install of Unraid and I could only get into C3 c-state. No matter what I did, referenced these for what to adjust in bios: https://www.reddit.com/r/unRAID/comments/1crs86u/thoughts_on_the_cwwk_h670_q670_board/ https://matthewhill.uk/general/cwwk-q670-low-power-intel-12-13-14-gen-nas-motherboard/

https://forum.level1techs.com/t/asrock-rack-w680d4id-2t-g5-x550-c-states-aspm-problem/221103/27

And also installed powertop and adjust all the tunables except the two Ethernet I226-V controller which causes the whole system to freeze if i enable them there and also disabled the wakeon stuff.

The KC3000 idled around 51-53c and If i enabled ASPM for both of them one of them always vanished from Unraid.

I then returned the KC3000 (2x 2TB) and swapped it with the Samsung 990 Pro (2x 2TB) without changing anything except being able to enable ASPM for both drives, and just like that the power consumption in idle dropped down to 21-22W and the idle temps are around 36c (I have a 7mm heatsink on the drives for both the KC3000 and the 990 Pro), however now the c-state is in C2...I will likely check try a few more settings in bios but i frankly dont expect it to help.

12 minutes ago, silicon4079 said:

Warning, avoid the Kingston KC3000 due to high power usage and high temps in idle!

Yeah, it's sadly valid for ALL enterprise SSDs. They don't support sleep states as consumer SSDs do.

9 hours ago, silicon4079 said:

The KC3000 idled around 51-53c

Not my experience:

Sorry, didn't know that Kingston has Consumer SSDs, too 😅

1 hour ago, JorgeB said:

Not my experience:

Interesting, which plugin gives you the wattage of individual devices?

6 minutes ago, googleg said:

which plugin gives you the wattage of individual devices?

It's a stock feature, you can enable in Settings - Disk Settings - Enable NVME power monitoring

5 hours ago, JorgeB said:

Not my experience:

What case are you using this in?

I sadly did not have a full screenshot but I do have this little one.

This is with the KC3000.

Also I highly doubt that power value is accurate at all seeing as when I disabled the pcie slots for the nvme when i had the KC3000 it dropped from 26-27W to 22-23W....certainly not 0.12W.

I dug a bit deeper and seems that enabling auto for ASPM for the NIC pcie root slot (1) was what was needed for it to have ASPM enabled without issues. Dropped idle to 20W from 21-22W. Still in C2...so I really have no clue what is causing this.

15 hours ago, mgutt said:

Yeah, it's sadly valid for ALL enterprise SSDs. They don't support sleep states as consumer SSDs do.

Is the Samsung 990 Pro a consumer SSD? Same price ish as the KC3000 so kinda doubt it...

Power values from unraid though.

Edited October 30, 2025Oct 30 by silicon4079

43 minutes ago, silicon4079 said:

What case are you using this in?

It's a Fractal Design something XL, don't remember the exact model.

The two cooler ones are on a PCIe adapter with better airflow from the side fan; the hotter one is in an M.2 slot.

44 minutes ago, silicon4079 said:

Also I highly doubt that power value is accurate at all

It may not be; it's what the devices themselves report, but it should be close, I have some Kioxia ones that report 0.01W at idle:

On a different server have a different Kingston model that doesn't report below 1.5W even at idle:

Just now, JorgeB said:

It's a Fractal Design something XL, don't remember the exact model.

Define 7 XL maybe or something that looks like that? That would explain it I guess. My nvme are on the backside on the mb and inside the Jonsbo N3 (ITX) case with barely any actual airflow down there I believe.

2 minutes ago, JorgeB said:

It may not be; it's what the devices themselves report

Have you checked with an external method like a smart power plug (which is what im using)?

8 minutes ago, silicon4079 said:

Have you checked with an external method like a smart power plug (which is what im using)?

I have a power meter, but there are two servers, a UPS, a router, and a switch all connected, so it's not easy to get small variations.

50 minutes ago, JorgeB said:

I have a power meter, but there are two servers, a UPS, a router, and a switch all connected, so it's not easy to get small variations.

Got it. The difference in temps at least is due to our case differences. As for power. i have no clue why my setup with the KC3000 used so much more than the 990 Pro im using now.

I swaped my i7-4770T with a Xeon 1275L V3 and now it draws 5 - 6 W more than before after running
powertop --auto-tune &>/dev/null

Does anyone know if the Adaptec ASR-78165 supports C-States or does this have the same issue the LSI cards have?

Edited November 19, 2025Nov 19 by jdiacobbo
Typo

I'm seeing some errors when I run Powertop on 7.2. Anyone else have similar errors and how to fix it? Searching didn't get me very far.

modprobe cpufreq_stats failed
Loaded 0 prior measurements
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 f
RAPL device for cpu 0
RAPL Using PowerCap Sysfs : Domain Mask f
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

Hi all,

I currently try to "eco" my server. For a couple of weeks I now try to find a way to lower the consumption. Unfortunately I don't reach any other state than C7. Using --auto-tune and stopping the array wont make it start again.

My Setup:

mini-ITX Z590I AORUS ULTRA

Platinum certified PSU

Intel® Core™ i5-10600 CPU @ 3.30GHz

16GB DDR4 RAM

6xHDD 1xSSD

MZHOU PCI-E 4x 6Port ASM1166

In the current state it drains in idle around 39-43 watts.

Not optimal but still better than the ~55 before..

I just today successfully flashed my ASM1166 card to the recommended firmware. Was hoping that this will do the med_power_with_dipm support trick, but unfortunately nothing has changed. As soon as I Enable SATA link power management for host*, I cannot restart the array.

00:01.0 PCI bridge: Intel Corporation 6th-10th Gen Core Processor PCIe Controller (x16) (rev 03) (prog-if 00 [Normal decode])
		LnkCap:	Port #2, Speed 8GT/s, Width x16, ASPM L0s L1, Exit Latency L0s <256ns, L1 <8us
		LnkCtl:	ASPM L0s L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
00:1b.0 PCI bridge: Intel Corporation Tiger Lake-H PCIe Root Port #17 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #17, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
		LnkCtl:	ASPM L0s L1 Enabled; RCB 64 bytes, LnkDisable- CommClk-
00:1b.4 PCI bridge: Intel Corporation Device 43c4 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #21, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
		LnkCtl:	ASPM L0s L1 Enabled; RCB 64 bytes, LnkDisable- CommClk-
00:1c.0 PCI bridge: Intel Corporation Tiger Lake-H PCIe Root Port #1 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #1, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
		LnkCtl:	ASPM L0s L1 Enabled; RCB 64 bytes, LnkDisable- CommClk-
00:1c.4 PCI bridge: Intel Corporation Tiger Lake-H PCI Express Root Port #5 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #5, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <16us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
00:1c.7 PCI bridge: Intel Corporation Device 43bf (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #8, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <16us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
00:1d.0 PCI bridge: Intel Corporation Tiger Lake-H PCI Express Root Port #9 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #9, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <16us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
00:1d.4 PCI bridge: Intel Corporation Device 43b4 (rev 11) (prog-if 00 [Normal decode])
		LnkCap:	Port #13, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s <1us, L1 <4us
		LnkCtl:	ASPM L0s L1 Enabled; RCB 64 bytes, LnkDisable- CommClk-
01:00.0 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, LnkDisable- CommClk+
05:00.0 Ethernet controller: Intel Corporation Ethernet Controller I225-V (rev 03)
		LnkCap:	Port #0, Speed 5GT/s, Width x1, ASPM L1, Exit Latency L1 <4us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
06:00.0 Network controller: Intel Corporation Wi-Fi 6 AX200 (rev 1a)
		LnkCap:	Port #0, Speed 5GT/s, Width x1, ASPM L1, Exit Latency L1 <8us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+
07:00.0 Non-Volatile memory controller: Sandisk Corp SanDisk Extreme Pro / WD Black 2018/SN750/PC SN720 NVMe SSD (prog-if 02 [NVM Express])
		LnkCap:	Port #0, Speed 8GT/s, Width x4, ASPM L1, Exit Latency L1 <8us
		LnkCtl:	ASPM L1 Enabled; RCB 64 bytes, LnkDisable- CommClk+

I dont know what these L[0-1] state's mean, but I guess L1 enabled means it is capable of power savings but stays at L0 (in use maybe?)

Edited November 22, 2025Nov 22 by W0nderW0lf

Can someone verify that my bios settings are alright, or have I missed something?

On 11/22/2025 at 9:38 AM, W0nderW0lf said:

Can someone verify that my bios settings are alright, or have I missed something?

You need to explicity set C State Limit to C10. "auto" often means "disabled".

On 11/21/2025 at 6:48 PM, W0nderW0lf said:

I cannot restart the array.

This is not normal. Check your syslog. The block diagram of the board looks pretty normal (not additional controllers visible). So maybe something related to the ASMedia card?!

@mgutt
Thank you very much for taking a look and ur detailed guide on this specific topic.

As you suggested "auto" really meaned "disabled" in this case.

I still need to check out why the array won't come online if I spin it down.
I will try that later.

So if I remove the Intel x710da2 I can reach c3, but if I install it, I only reach c2. Why is this and how can I prevent it? ASPM accepted/enabled.

On 11/25/2025 at 5:45 AM, Nozle said:

So if I remove the Intel x710da2 I can reach c3, but if I install it, I only reach c2. Why is this and how can I prevent it? ASPM accepted/enabled.

Not an answer, but I reach C3 with an x710da2.

I have yet to determine why I can't do better than C3... Asus WS680-ACE IPMI here (and I suspect IPMI may be the blocker for me - had a USB device forcing C2)

i was just wondering..

are any of you guys using all the arrs while trying to maintain a as low as possible power usage?

my idle for my i5-11600 with 32gb memory and 5x 3,5 inch and 7 ssd/nvme doesnt go below 60 to 65 watt..

i cant power down disks though, it will fire them up constantly due to plex/arrs..

On 5/10/2025 at 9:02 PM, NCM said:

RC6pp       0.0%    |

See if you can enable these.

GT/Media - Power Management Control:

RC6 (Render Standby) disabled to Enabled

MC6 (Media Standby) disabled to Enabled

It is in the advanced tab then inside power & performance page for me.

With my Intel Ultra 125H they were disabled in the bios and enabling them dropped my idle from around 15w to 10w. This is a NUC with 96GB ram and 2TB NVME running proxmox.

17 hours ago, furian said:

5x 3,5 inch and 7 ssd/nvme doesnt go below 60 to 65 watt..

have you looked up standby power consumption for those?

It's 5x spinning rust + 7x ssd so 12 physical devices.

There is a limit to what kind of power consumption you can save from standby and spindown.

however what you are describing sounds like something is constantly writing to your array.

do you have apps on a separate nvme/ssd only pool? If not there's your answer as to why the hdds constantly spin up.

Edited December 6, 2025Dec 6 by jit-010101