12gen N-series Nas motherboard (topton, cwwk, ... )

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

Vitobambino

New Member
Aug 18, 2024
12
1
3
I received my CW-ALDN-NAS-K (the purple N305) last week. I'm using it mainly for my Unraid NAS to replace a 10-year-old i3-3340 due to its lack of hardware decoding ability and high power consumption.

After two full days of tuning, Googling, and reading tech articles, I still think it's a good MOBO. If you're interested in diving in, here are some things you need to know:
  1. RAM Compatibility: This MOBO is very picky about RAM. Micron Crucial, Samsung, and SK Hynix are good choices. Don't buy RAM with a frequency higher than 4800MHz; it's not supported by the CPU.
  2. USB Ports: According to comments from beta testers on the official page, the bottom port of the onboard USB is unavailable due to the lack of a USB 2.0 port from the CPU. Only the upper one is working.
  3. M.2 Slot Spacing: The two M.2 slots are too close to each other, so the heatsink cannot be wider than your NVMe drive. Don't buy M.2 coolers, as most of them are wider than the M.2 itself and even larger than the M.2 heatsink.
  4. CPU Contact Issue: The CPU doesn't make good contact with the copper plate. Prepare a 0.5mm thermal pad to fix it.

Here's my setup:
View attachment 39628
Code:
MOBO: CW-ALDN-NAS-K
RAM:  Crucial CT32G48C40S5 32GB 4800MHz
M.2:  KLEVV CRAS CV710 512GB x2
HDD:  10TB x1, 3TB x1, 1TB x1
OS:   unRAID 6.12.13

BIOS Configuration (Press Ctrl+F1 to open hidden options):
Markdown (GitHub flavored):
Advanced
- Power & Performance
  - CPU
    - C states
      - Package C State Limit: C10
- RC ACPI Settings
  - Native PCIE Enable: Enabled
  - Native ASPM: Enabled

Chipset
- System Agent (SA) Configuration
  - Memory Configuration
    - In-Band ECC Support: Enabled # optional
  - Graphics Configuration
    - Skip Scaning of External Gfx Card: Enabled # optional
    - RC1p Support: Enabled
  - TCSS setup menu
    - D3 Cold Enable/Disable: Disabled # prevent it died after sleep
  - DMI/OPI Configuration
    - DMI Gen3 ASPM: ASPM L0sL1
    - DMI ASPM: ASPM L0sL1
- PCH-IO Configuration
  - PCI Express Configuration
    - DMI Link ASPM Control: L0sL1
    - PCI Express Root Port {1,2,3,4,7,9}
      - ASPM: L1
      - L1 Substates: L1.1 & L1.2
      - L1 Low: Enabled
  - SATA Configuration
    - SATA Controller(s): Disabled # this mobo use ASM1166, not native one
  - HD Audio Subsystem Configuration Settings
    - HD Audio: Disabled # optional
  - SCS Configuration
    - eMMC 5.1 Controller: Disabled # optional

Startup Script:
Bash:
# -------------------------------------------------
# Set power-efficient CPU governor
# -------------------------------------------------
/etc/rc.d/rc.cpufreq powersave

# -------------------------------------------------
# Disable CPU Turbo
# -------------------------------------------------
[[ -f /sys/devices/system/cpu/intel_pstate/no_turbo ]] && echo "1" > /sys/devices/system/cpu/intel_pstate/no_turbo
[[ -f /sys/devices/system/cpu/cpufreq/boost ]] && echo "0" > /sys/devices/system/cpu/cpufreq/boost

# -------------------------------------------------
# Powertop Tweaks
# -------------------------------------------------

# Disable Wake-on-LAN
for i in /sys/class/net/eth?; do ethtool -s $(basename $i) wol d; done

# Runtime PM for I2C Adapter (i915 gmbus dpb)
echo auto | tee /sys/bus/i2c/devices/i2c-*/device/power/control

# Autosuspend for USB devices
echo auto | tee /sys/bus/usb/devices/*/power/control

# Runtime PM for disk
echo auto | tee /sys/block/sd*/device/power/control

# Host bridge
echo 'auto' > '/sys/bus/pci/devices/0000:00:00.0/power/control';

# RAM memory: Intel Corporation Alder Lake-N PCH Shared SRAM
echo 'auto' > '/sys/bus/pci/devices/0000:00:14.2/power/control';

# ISA bridge: Intel Corporation Alder Lake-N PCH eSPI Controller
echo 'auto' > '/sys/bus/pci/devices/0000:00:1f.0/power/control';

# Serial bus controller: Intel Corporation Alder Lake-N SPI (flash) Controller
echo 'auto' > '/sys/bus/pci/devices/0000:00:1f.5/power/control';

# NVME
echo 'auto' > '/sys/bus/pci/devices/0000:01:00.0/power/control';

# NVME
echo 'auto' > '/sys/bus/pci/devices/0000:02:00.0/power/control';

# SATA controller: ASMedia Technology Inc. ASM1166 Serial ATA Controller (rev 02)
echo 'auto' > '/sys/bus/pci/devices/0000:03:00.0/power/control';

# Runtime PM for ATA devices
echo auto | tee /sys/bus/pci/devices/0000:03:00.0/ata*/power/control

With these configurations, the CPU can enter C8 state:
Code:
           Pkg(HW)  |            Core(HW) |            CPU(OS) 0
                    |                     | C0 active   4.6%
                    |                     | POLL        0.0%    0.0 ms
                    |                     | C1E         4.7%    0.1 ms
C2 (pc2)   14.7%    |                     |
C3 (pc3)    0.8%    | C3 (cc3)    0.0%    |
C6 (pc6)   17.3%    | C6 (cc6)   85.7%    | C6          7.0%    0.9 ms
C7 (pc7)    0.0%    | C7 (cc7)   79.7%    |
C8 (pc8)   13.3%    |                     | C8          3.6%    1.1 ms
C9 (pc9)    0.0%    |                     |
C10 (pc10)  0.0%    |                     |
                    |                     | C10        78.6%    6.2 ms

Powertop Output:
Code:
>> Bad           Enable SATA link power management for host4
   Bad           Enable SATA link power management for host5
   Bad           Enable SATA link power management for host3
   Bad           Enable SATA link power management for host1
   Bad           Enable SATA link power management for host6
   Bad           Enable SATA link power management for host2
   Bad           VM writeback timeout
   Bad           Runtime PM for PCI Device Intel Corporation Ethernet Controller I226-V
   Bad           Runtime PM for PCI Device Intel Corporation Ethernet Controller I226-V
There are still some "BAD"s:
  1. SATA Link Power Management: Not supported by the ASM1166 controller.
  2. Intel I226-V Ethernet Controller: Has a power management issue. Enabling it causes serious OS hangs (though there's no issue if running on Arch Linux with the latest Linux kernel).
  3. VM writeback timeout: Can be ignored if you have UPS.
This configuration has been running for 24 hours without problems, even under full usage scenarios:
  • Jellyfin: Software decoding of high-bitrate H.264 High 10 video (no hardware decoding available) combined with HEVC hardware encoding. This results in full CPU and GPU loading.

ref:
  1. Reduce power consumption with powertop - unRAD
  2. The Curse of ASPM
Great manual, thanks.
What power consumption you achieved after these tweaks (Bios settings and startup script)?
 
  • Like
Reactions: Zorg2K

Zorg2K

New Member
Aug 10, 2024
15
5
3
Belgium
FYI Here are my temps and consumption

For the Purple N305 in an N2 Case with the stock Jonsbo rad+fan, the CPU Package temp is at 33°C doing a disk preclear, at 30% load
and 30°C at 2% (Currently only Pi-Hole-Unbound and JellyFin running).

It idles at ~24w with disks spun down and no other optimization.

I also bought a thermostat PWM Fan controller so the case fan isn't too noisy and uses power only when needed. I works great !

DC8-16V 4 Wire PWM PC CPU Fan Temperature Control Thermostat Speed Controller.png


~Zorg2K
 

Zorg2K

New Member
Aug 10, 2024
15
5
3
Belgium
@MiauLightouch

Thank you so much for this. Ideal timing :)

Two small questions though :

  • Which Powertop Version did you install on UnRaid? 2.15?
  • If you completely disable the boost, does it mean you run at the base frequency of 0,8GHz ?

Code:
# -------------------------------------------------
# Disable CPU Turbo
# -------------------------------------------------
[[ -f /sys/devices/system/cpu/intel_pstate/no_turbo ]] && echo "1" > /sys/devices/system/cpu/intel_pstate/no_turbo
[[ -f /sys/devices/system/cpu/cpufreq/boost ]] && echo "0" > /sys/devices/system/cpu/cpufreq/boost
 

Outbound0918

Member
Jun 29, 2024
46
18
8
@MiauLightouch, Great write up to consolidate everything! :)

I have a few points I would like to comment on however.

RAM Compatibility: This MOBO is very picky about RAM. Micron Crucial, Samsung, and SK Hynix are good choices. Don't buy RAM with a frequency higher than 4800MHz; it's not supported by the CPU.
I agree on the Memory Manufacturers, SK Hynix is working fine for me. However, it does not matter if you use a module with a frequency > 4800MHz as it will just be downclocked to what the CPU supports. For exmaple a 5600MHz module is fine to use and will just be automatically downclocked to 4800Mhz.

USB Ports: According to comments from beta testers on the official page, the bottom port of the onboard USB is unavailable due to the lack of a USB 2.0 port from the CPU. Only the upper one is working.
I have zero issues with any of the USB ports, can you explain further?

BIOS Configuration (Press Ctrl+F1 to open hidden options):
If you have to do this, you are on an old BIOs version, try upgrading and re-testing your findings.

# ------------------------------------------------- # Disable CPU Turbo # ------------------------------------------------- [[ -f /sys/devices/system/cpu/intel_pstate/no_turbo ]] && echo "1" > /sys/devices/system/cpu/intel_pstate/no_turbo [[ -f /sys/devices/system/cpu/cpufreq/boost ]] && echo "0" > /sys/devices/system/cpu/cpufreq/boost
This seems odd to want to disbale Turbo Boosting. Your CPU will function at a lower clock when doing tasks but the tasks will take longer so it is better to keep Turbo Boosting enabled so the tasks are completed more quickly at a higher clock. Completing the tasks quickly at higher clock compared to completing the tasks slowly at a lower clock should have comparable power consumption.

With these configurations, the CPU can enter C8 state:
My system is actually entering C10 state. I have only enabled L1 ASPM state wherever I am able to and did not enable L0s or L0sL1 anywhere.

SATA Link Power Management: Not supported by the ASM1166 controller.
The ASM1166 controller 100% supports SATA Link Power Management. The ASM spec even states it supports Partial/Slumber power management, which means SATA Link Power Management IS supported. Maybe you need to update your ASM1166 to a newer version.

Intel I226-V Ethernet Controller: Has a power management issue. Enabling it causes serious OS hangs (though there's no issue if running on Arch Linux with the latest Linux kernel).
How are you testing this, what are you enabling and disabling exactly and where?
I want to verify this on my system.

VM writeback timeout: Can be ignored if you have UPS.
Please expand on this? VM writeback timeout does not mean much or explain the problem you are having.
 
Last edited:
  • Like
Reactions: Zorg2K

Zorg2K

New Member
Aug 10, 2024
15
5
3
Belgium
@Outbound0918 @MiauLightouch

I agree on the Memory Manufacturers, SK Hynix is working fine for me. However, it does not matter if you use a module with a frequency > 4800MHz as it will just be downclocked to what the CPU supports. For exmaple a 5600MHz module is fine to use and will just be automatically downclocked to 4800Mhz.
Confirmed my 5600MHz Crucial CT32G56C46S5 works like a charm, I could have saved 30€ though.

If you have to do this, you are on an old BIOs version, try upgrading and re-testing your findings.
I have (I think) the latest BIOS installed (CW-NAS-ADLN-KV10_BootSATA_20240624.iso) and still need to do CTRL-F1 for Advanced Settings.

This seems odd to want to disable Turbo Boosting. Your CPU will function at a lower clock when doing tasks but the tasks will take longer so it is better to keep Turbo Boosting enabled so the tasks are completed more quickly at a higher clock. Completing the tasks quickly at higher clock compared to completing the tasks slowly at a lower clock should have comparable power consumption.
Yeah. Had the same thought. Not sure about the different denominations though. I assume Turbo Boost, Turbo Speed, etc. are the same. But why remove this when the CPU is especially designed to step down?
 
  • Like
Reactions: Outbound0918

MiauLightouch

New Member
Oct 15, 2024
5
9
3
I agree on the Memory Manufacturers, SK Hynix is working fine for me. However, it does not matter if you use a module with a frequency > 4800MHz as it will just be downclocked to what the CPU supports. For exmaple a 5600MHz module is fine to use and will just be automatically downclocked to 4800Mhz.
That's what I meant. You can buy any module higher than 4800MHz, but it would be a waste since it will just be downclocked. I will modify my wording; thank you for your suggestion.

---

I have zero issues with any of the USB ports, can you explain further?
The USB port beside the PCIe lane has two ports. The bottom port is non-functional, at least on my board.

---

If you have to do this, you are on an old BIOs version, try upgrading and re-testing your findings.
There's no HD Audio option unless you open the hidden menu.

---

This seems odd to want to disbale Turbo Boosting. Your CPU will function at a lower clock when doing tasks but the tasks will take longer so it is better to keep Turbo Boosting enabled so the tasks are completed more quickly at a higher clock. Completing the tasks quickly at higher clock compared to completing the tasks slowly at a lower clock should have comparable power consumption.
Yes, that makes sense. I will adjust my post accordingly. And don't forget, I'm just sharing my config. My workload is not that heavy to require enabling CPU boost, and it actually takes more power—from idle 15W to 38W. Feel free to ignore the boost config.

---

My system is actually entering C10 state. I have only enabled L1 ASPM state wherever I am able to and did not enable L0s or L0sL1 anywhere.
Could you please share your configuration?

---

The ASM1166 controller 100% supports SATA Link Power Management. The ASM spec even states it supports Partial/Slumber power management, which means SATA Link Power Management IS supported. Maybe you need to update your ASM1166 to a newer version.
If I haven't misread, according to this [post](SATA Controller part), it mentions there is incompatibility when setting the link power management policy.

---

How are you testing this, what are you enabling and disabling exactly and where?
I want to verify this on my system.
  1. I performed a fresh install of Arch Linux and ran `powertop --auto-tune`.
  2. In my current Unraid instance, I manually enabled it in the boot script or via `powertop` (in the 'Tunables' tab).
---

Please expand on this? VM writeback timeout does not mean much or explain the problem you are having.
This "VM writeback timeout" issue is explained in detail in this forum post: [Reduce power consumption with powertop].
 
Last edited:

Outbound0918

Member
Jun 29, 2024
46
18
8
Could you please share your configuration?
Sure, will grab this for you.

If I haven't misread, according to this [post](SATA Controller part), it mentions there is incompatibility when setting the link power management policy.
No no, the ASM1166 **does** support Link Power Management. There was a bug in early firmware versions that prevented it from working, but in the latest version it works fine.

This "VM writeback timeout" issue is explained in detail in this forum post: [Reduce power consumption with powertop].
This sounds odd, I will probably read up on this to understand better.
 

Zorg2K

New Member
Aug 10, 2024
15
5
3
Belgium
@MiauLightouch

I tried your suggested settings but without disabling Turbo Boost. As one would expect, I don't reach C8 state. But the other tweaks saved me a nice ~2.5W without crashing, a welcome improvement since powertop --auto-tune crashes the UnRAID badly in no time. That's a 8€ saving a year ! So thank you for this

Code:
           Pkg(HW)  |            Core(HW) |            CPU(OS) 0
                    |                     | C0 active   1.6%
                    |                     | POLL        0.0%    0.0 ms
                    |                     | C1E         2.0%    0.4 ms
C2 (pc2)    1.7%    |                     |
C3 (pc3)   65.0%    | C3 (cc3)    0.0%    |
C6 (pc6)    0.0%    | C6 (cc6)   94.9%    | C6          7.8%    0.9 ms
C7 (pc7)    0.0%    | C7 (cc7)   87.5%    |
C8 (pc8)    0.0%    |                     | C8          0.9%    2.4 ms
C9 (pc9)    0.0%    |                     |
C10 (pc10)  0.0%    |                     |
                    |                     | C10        87.4%   14.1 ms
 
Last edited:

MiauLightouch

New Member
Oct 15, 2024
5
9
3
No no, the ASM1166 **does** support Link Power Management. There was a bug in early firmware versions that prevented it from working, but in the latest version it works fine.
I'm extremely happy to hear you confirm this; it's news to me. I think I was too afraid to enable this option due to concerns about data loss. I'll try it on my Unraid later. Thank you for pointing out the error.
 

Outbound0918

Member
Jun 29, 2024
46
18
8

Outbound0918

Member
Jun 29, 2024
46
18
8
I received my CW-ALDN-NAS-K (the purple N305) last week. I'm using it mainly for my Unraid NAS to replace a 10-year-old i3-3340 due to its lack of hardware decoding ability and high power consumption.

After two full days of tuning, Googling, and reading tech articles, I still think it's a good MOBO. If you're interested in diving in, here are some things you need to know:
  1. RAM Compatibility: This MOBO is very picky about RAM. Micron Crucial, Samsung, and SK Hynix are good choices. Don't buy RAM with a frequency higher than 4800MHz; it's not supported by the CPU. update: it will be automatically downclocked to 4800MHz.
  2. USB Ports: According to comments from beta testers on the official page, the bottom port of the onboard USB (update: beside the PCIe Lane) is unavailable due to the lack of a USB 2.0 port from the CPU. Only the upper one is working.
  3. M.2 Slot Spacing: The two M.2 slots are too close to each other, so the heatsink cannot be wider than your NVMe drive. Don't buy M.2 coolers, as most of them are wider than the M.2 itself and even larger than the M.2 heatsink.
  4. CPU Contact Issue: The CPU doesn't make good contact with the copper plate. Prepare a 0.5mm thermal pad to fix it.

Here's my setup:
View attachment 39628
Code:
MOBO: CW-ALDN-NAS-K
RAM:  Crucial CT32G48C40S5 32GB 4800MHz
M.2:  KLEVV CRAS CV710 512GB x2
HDD:  10TB x1, 3TB x1, 1TB x1
OS:   unRAID 6.12.13

BIOS Configuration (Press Ctrl+F1 to open hidden options):
Markdown (GitHub flavored):
Advanced
- Power & Performance
  - CPU
    - C states
      - Package C State Limit: C10
- RC ACPI Settings
  - Native PCIE Enable: Enabled
  - Native ASPM: Enabled

Chipset
- System Agent (SA) Configuration
  - Memory Configuration
    - In-Band ECC Support: Enabled # optional
  - Graphics Configuration
    - Skip Scaning of External Gfx Card: Enabled # optional
    - RC1p Support: Enabled
  - TCSS setup menu # optional: fix thunderbolt errors
    - D3 Cold Enable/Disable: Disabled
    - D3Hot: Disabled
  - DMI/OPI Configuration
    - DMI Gen3 ASPM: ASPM L1
    - DMI ASPM: ASPM L1
- PCH-IO Configuration
  - PCI Express Configuration
    - DMI Link ASPM Control: L1
    - PCI Express Root Port {1,2,3,4,7,9}
      - ASPM: L1
      - L1 Substates: L1.1 & L1.2
      - L1 Low: Enabled
  - SATA Configuration
    - SATA Controller(s): Disabled # this mobo use ASM1166, not native one
  - HD Audio Subsystem Configuration Settings
    - HD Audio: Disabled # optional
  - SCS Configuration
    - eMMC 5.1 Controller: Disabled # optional

Startup Script:
Bash:
# -------------------------------------------------
# Set power-efficient CPU governor
# -------------------------------------------------
/etc/rc.d/rc.cpufreq powersave

# -------------------------------------------------
# Disable CPU Turbo
# -------------------------------------------------

# update: optional, uncomment this part if you don't need CPU Boost
#[[ -f /sys/devices/system/cpu/intel_pstate/no_turbo ]] && echo "1" > /sys/devices/system/cpu/intel_pstate/no_turbo
#[[ -f /sys/devices/system/cpu/cpufreq/boost ]] && echo "0" > /sys/devices/system/cpu/cpufreq/boost

# -------------------------------------------------
# Enable power-efficient ethernet
# -------------------------------------------------

# Disable Wake-on-LAN
for i in /sys/class/net/eth?; do ethtool -s $(basename $i) wol d; done

# -------------------------------------------------
# Powertop Tweaks
# -------------------------------------------------

# Enable SATA link power management
echo med_power_with_dipm | tee /sys/class/scsi_host/host*/link_power_management_policy

# Runtime PM for I2C Adapter (i915 gmbus dpb)
echo auto | tee /sys/bus/i2c/devices/i2c-*/device/power/control

# Autosuspend for USB devices
echo auto | tee /sys/bus/usb/devices/*/power/control

# Runtime PM for disk
echo auto | tee /sys/block/sd*/device/power/control

# Host bridge
echo auto | tee /sys/bus/pci/devices/0000:00:00.0/power/control

# RAM memory: Intel Corporation Alder Lake-N PCH Shared SRAM
echo auto | tee /sys/bus/pci/devices/0000:00:14.2/power/control

# ISA bridge: Intel Corporation Alder Lake-N PCH eSPI Controller
echo auto | tee /sys/bus/pci/devices/0000:00:1f.0/power/control

# Serial bus controller: Intel Corporation Alder Lake-N SPI (flash) Controller
echo auto | tee /sys/bus/pci/devices/0000:00:1f.5/power/control

# -------------------------------------------------
# execute `lspci -tv` to fit your build
# -------------------------------------------------

# NVME
echo auto | tee /sys/bus/pci/devices/0000:01:00.0/power/control

# NVME
echo auto | tee /sys/bus/pci/devices/0000:02:00.0/power/control

# SATA controller: ASMedia Technology Inc. ASM1166 Serial ATA Controller (rev 02)
echo auto | tee /sys/bus/pci/devices/0000:03:00.0/power/control

# Runtime PM for ATA devices
echo auto | tee /sys/bus/pci/devices/0000:03:00.0/ata*/power/control

# Ethernet controller: Intel Corporation Ethernet Controller I226-V (rev 04)
# DO NOT uncomment; it may cause serious OS hangs, especially when opening powertop
#echo auto | tee /sys/bus/pci/devices/0000:04:00.0/power/control
#echo auto | tee /sys/bus/pci/devices/0000:05:00.0/power/control

With these configurations, the CPU can enter C8 state:
Code:
           Pkg(HW)  |            Core(HW) |            CPU(OS) 0
                    |                     | C0 active   4.6%
                    |                     | POLL        0.0%    0.0 ms
                    |                     | C1E         4.7%    0.1 ms
C2 (pc2)   14.7%    |                     |
C3 (pc3)    0.8%    | C3 (cc3)    0.0%    |
C6 (pc6)   17.3%    | C6 (cc6)   85.7%    | C6          7.0%    0.9 ms
C7 (pc7)    0.0%    | C7 (cc7)   79.7%    |
C8 (pc8)   13.3%    |                     | C8          3.6%    1.1 ms
C9 (pc9)    0.0%    |                     |
C10 (pc10)  0.0%    |                     |
                    |                     | C10        78.6%    6.2 ms

Powertop Output:
Code:
>> Bad           VM writeback timeout
   Bad           Runtime PM for PCI Device Intel Corporation Ethernet Controller I226-V
   Bad           Runtime PM for PCI Device Intel Corporation Ethernet Controller I226-V
There are still some "BAD"s:
  1. SATA Link Power Management: Not supported by the ASM1166 controller.
  2. Intel I226-V Ethernet Controller: Has a power management issue. Enabling it causes serious OS hangs (though there's no issue if running on Arch Linux with the latest Linux kernel).
  3. VM writeback timeout: Can be ignored if you have UPS.
This configuration has been running for 24 hours without problems, even under full usage scenarios:
  • Jellyfin: Software decoding of high-bitrate H.264 High 10 video (no hardware decoding available) combined with HEVC hardware encoding. This results in full CPU and GPU loading.

update: Special thanks to @Outbound0918 for the valuable suggestions and guidance, which helped me resolve many issues.

ref:
  1. Reduce power consumption with powertop - unRAD
  2. The Curse of ASPM
I just spent some time going through this list in my BIOs and they were pretty similar to yours, but I will list my comments below.

Markdown (GitHub flavored):
- PCH-IO Configuration
  - PCI Express Configuration
    - DMI Link ASPM Control: L1
    - PCI Express Root Port {1,2,3,4,7,9}
      - ASPM: L1
      - L1 Substates: L1.1 & L1.2
      - L1 Low: Enabled
There are 9 PCIE lanes in total, this only configures ASPM for 6 of them, the other 3 are under: System Agent (SA) Configuration -> PCI Express Configuration -> Ports-> "ASPM stuff"
You should configure ASPM for the lanes here too.

Markdown (GitHub flavored):
- Power & Performance
  - CPU
    - C states
      - Package C State Limit: C10
The OS seems to override this for me, most likely set in some kernel parameters.

Markdown (GitHub flavored):
- System Agent (SA) Configuration
You can also enable the setting C10 Dynamic Threshold Adjustment. I have not tested if it improves anything in terms of power savings, but it *theoretically should*.

Markdown (GitHub flavored):
  - TCSS setup menu # optional: fix thunderbolt errors
    - D3 Cold Enable/Disable: Disabled
    - D3Hot: Disabled
Where are you seeing these thunderbolt errors and what issues are they causing you?


Markdown (GitHub flavored):
  - Memory Configuration
    - In-Band ECC Support: Enabled # optional
Some notes for others on these settings:
In-Band ECC Support: Enabled # Allows error-correcting functionality on non-ECC memory by using part of the system's regular memory for error correction.
ECC operation mode: 2 # All memory requests are protected by ECC
IBECC error injection control: no error injection # Used for testing if in-band ecc works, make sure it is "no error injection" for normal use.

The setting below in **BOLD** disable themselves if you go back into the BIOs after saving. They are in fact still enabled if you DO NOT re-save when they show as disabled.
What I am doing is configuring the BIOs, saving to User Defaults, then saving and exiting. Then if I want to make anymore BIOs changes, restore user defaults, make changes, save to user defaults, then save and exit. Annoying yes, but it is the best workaround until this bug is fixed by CWWK.
Markdown (GitHub flavored):
- PCH-IO Configuration
  - PCI Express Configuration
    - DMI Link ASPM Control: L1
    - PCI Express Root Port {1,2,3,4,7,9}
[B]      - ASPM: L1 # This gets disabled after reboot
      - L1 Substates: L1.1 & L1.2 # This gets disabled after reboot[/B]
 
Last edited:

Outbound0918

Member
Jun 29, 2024
46
18
8
# ------------------------------------------------- # Enable power-efficient ethernet # ------------------------------------------------- # Disable Wake-on-LAN for i in /sys/class/net/eth?; do ethtool -s $(basename $i) wol d; done
Additionally, I highly doubt disabling WOL will save any power when the system is powered on.
 

MiauLightouch

New Member
Oct 15, 2024
5
9
3
There are 9 PCIE lanes in total, this only configures ASPM for 6 of them, the other 3 are under: System Agent (SA) Configuration -> PCI Express Configuration -> Ports-> "ASPM stuff"
You should configure ASPM for the lanes here too.
It's enabled to ASPM L1 by default.

PCIe port map under PCH-IO should be:
- Port 1: NVMe
- Port 2: NVMe
- Port 3: ASM1166
- Port 4: I226-V
- Port 7: I226-V
- Port 9: (Maybe PCIe x4 slot)

The OS seems to override this for me, most likely set in some kernel parameters.

You can also enable the setting C10 Dynamic Threshold Adjustment. I have not tested if it improves anything in terms of power savings, but it *theoretically should*.
I can confirm it's handled by `intel_idle` kernel module; there's no need to set it in the BIOS.

Where are you seeing these thunderbolt errors and what issues are they causing you?
Just suppressing error messages from dmesg.

The setting below in **BOLD** disable themselves if you go back into the BIOs after saving. They are in fact still enabled if you DO NOT re-save when they show as disabled.
What I am doing is configuring the BIOs, saving to User Defaults, then saving and exiting. Then if I want to make anymore BIOs changes, restore user defaults, make changes, save to user defaults, then save and exit. Annoying yes, but it is the best workaround until this bug is fixed by CWWK.
Markdown (GitHub flavored):
- PCH-IO Configuration
  - PCI Express Configuration
    - DMI Link ASPM Control: L1
    - PCI Express Root Port {1,2,3,4,7,9}
[B]      - ASPM: L1 # This gets disabled after reboot
      - L1 Substates: L1.1 & L1.2 # This gets disabled after reboot[/B]
I will add this section to my post; I forgot to mention it earlier.

Additionally, I highly doubt disabling WOL will save any power when the system is powered on.
It can be ignored I think.
 
  • Like
Reactions: Outbound0918

Outbound0918

Member
Jun 29, 2024
46
18
8
I can confirm it's handled by `intel_idle` kernel module; there's no need to set it in the BIOS.
"Influenced", not handled. ☺
Still should be enabled in the BIOs.

It's enabled to ASPM L1 by default.

PCIe port map under PCH-IO should be:
- Port 1: NVMe
- Port 2: NVMe
- Port 3: ASM1166
- Port 4: I226-V
- Port 7: I226-V
- Port 9: (Maybe PCIe x4 slot)
Then what are ports 5, 6 and 8 under: System Agent (SA) Configuration -> PCI Express Configuration -> Ports-> "ASPM stuff"
 

MiauLightouch

New Member
Oct 15, 2024
5
9
3
"Influenced", not handled. ☺
Still should be enabled in the BIOs.
Yes, you're right. It turns out I forgot that my user defaults set the cstate limit to C10.

Then what are ports 5, 6 and 8 under: System Agent (SA) Configuration -> PCI Express Configuration -> Ports-> "ASPM stuff"
N305 only has 9 PCIe lanes, and they're all under PCH-IO configuration.

We can leave the PCIe Ports under "System Agent (SA) Configuration" disabled; there's no difference.
 
  • Like
Reactions: Outbound0918

Wallie

New Member
Feb 18, 2024
7
3
3
On my BKHD-1264-NAS, I recently added an ASM1166 SATA adapter in the first M.2 slot. To do this, I had to move the NVMe SSD previously in that slot to a PCIe x1 adapter.

Has anyone found a useful purpose for the second slot?
 

Attachments

  • Like
Reactions: MurkyDudek

Vitobambino

New Member
Aug 18, 2024
12
1
3
Which Intel stock radiator fit to this board? (Purple n100 version). I ordered version with only copper plate. Should I remove it before mount radiator?
 
Last edited: