Poor power efficiency with B660/B760* boards and active** Full-Speed USB devices

[Dan0987]

Hi,

The CPU spends a significant amount of time in higher C-states after connecting a Full-Speed (12Mbit/s) USB device, significantly affecting power consumption. The deeper the Package C-state was in prior to connecting the Full-Speed USB device, the greater the difference in power consumption.

* the problem has been observed on motherboards with B660 and B760 chipsets, but it is possible that other 600 and 700 series chipsets are also affected.

** applies to connected and not suspended devices

Here are some numbers produced on a Linux system with an i5-13500 and a Gigabyte B660 AORUS MASTER DDR4 and various USB devices, but it appears that MS Windows is also affected:

Baseline. System at idle. Total power consumption: ~ 5.3 W

C2 (pc2)	0,6%
C3 (pc3)	0,7%
C6 (pc6)	0,3%
C7 (pc7)	0,0%
C8 (pc8)	0,0%
C9 (pc9)	0,0%
C10 (pc10)	97,5%

Baseline + Logitech M90 mouse (Low-Speed, 1.5 Mbit/s). Total power consumption: ~ 6.5 W

T:  Bus=01 Lev=01 Prnt=02 Port=06 Cnt=01 Dev#= 18 Spd=1.5  MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs=  1
P:  Vendor=046d ProdID=c05a Rev=63.00
S:  Manufacturer=Logitech
S:  Product=USB Optical Mouse
C:  #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=100mA
I:  If#= 0 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=02 Driver=usbhid
E:  Ad=81(I) Atr=03(Int.) MxPS=   4 Ivl=10ms

C2 (pc2)	5,0%
C3 (pc3)	1,1%
C6 (pc6)	0,3%
C7 (pc7)	0,0%
C8 (pc8)	0,0%
C9 (pc9)	0,0%
C10 (pc10)	92,7%

Baseline + Flash Drive (High-Speed, 480 Mbit/s). Total power consumption: ~ 6 W

T:  Bus=01 Lev=01 Prnt=02 Port=06 Cnt=01 Dev#= 16 Spd=480  MxCh= 0
D:  Ver= 2.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=0781 ProdID=5583 Rev=01.00
S:  Manufacturer=SanDisk
S:  Product=Ultra Fit
S:  SerialNumber=<redacted>
C:  #Ifs= 1 Cfg#= 1 Atr=80 MxPwr=224mA
I:  If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms

C2 (pc2)	0,8%
C3 (pc3)	1,0%
C6 (pc6)	0,3%
C7 (pc7)	0,0%
C8 (pc8)	0,0%
C9 (pc9)	0,0%
C10 (pc10)	96,9%

Baseline + spinning 5 TB 2,5" HDD powered from the USB port. (SuperSpeed, 5 Gbit/s). Total power consumption: ~ 8.6 W

T:  Bus=02 Lev=01 Prnt=01 Port=02 Cnt=01 Dev#=  4 Spd=5000 MxCh= 0
D:  Ver= 3.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs=  1
P:  Vendor=1058 ProdID=2665 Rev=20.07
S:  Manufacturer=Western Digital
S:  Product=My Passport 2665
S:  SerialNumber=<redacted>
C:  #Ifs= 1 Cfg#= 1 Atr=80 MxPwr=896mA
I:  If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage
E:  Ad=02(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=81(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms

C2 (pc2)	0,8%
C3 (pc3)	1,0%
C6 (pc6)	0,2%
C7 (pc7)	0,0%
C8 (pc8)	0,0%
C9 (pc9)	0,0%
C10 (pc10)	97,1%

And now, for comparison, some Full-Speed (12 Mbit/s) USB devices:

Baseline + Keyboard. Total power consumption: ~ 11.5 W

T:  Bus=01 Lev=01 Prnt=18 Port=07 Cnt=01 Dev#= 17 Spd=12   MxCh= 0
D:  Ver= 1.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=04d9 ProdID=a131 Rev=01.04
S:  Manufacturer=HOLTEK
S:  Product=USB-HID Keyboard
C:  #Ifs= 4 Cfg#= 1 Atr=a0 MxPwr=500mA
I:  If#= 0 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=01 Driver=usbhid
E:  Ad=81(I) Atr=03(Int.) MxPS=   8 Ivl=1ms
I:  If#= 1 Alt= 0 #EPs= 2 Cls=03(HID  ) Sub=00 Prot=00 Driver=usbhid
E:  Ad=04(O) Atr=03(Int.) MxPS=  64 Ivl=1ms
E:  Ad=83(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 2 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=00 Prot=00 Driver=usbhid
E:  Ad=82(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 3 Alt= 0 #EPs= 2 Cls=03(HID  ) Sub=00 Prot=00 Driver=usbhid
E:  Ad=06(O) Atr=03(Int.) MxPS=  64 Ivl=1ms
E:  Ad=85(I) Atr=03(Int.) MxPS=  64 Ivl=1ms

C2 (pc2)	51,4%
C3 (pc3)	1,1%
C6 (pc6)	0,1%
C7 (pc7)	0,0%
C8 (pc8)	0,1%
C9 (pc9)	0,0%
C10 (pc10)	46,2%

Baseline + Logitech G700 mouse. Total power consumption: ~ 11.5 W

T:  Bus=01 Lev=01 Prnt=02 Port=06 Cnt=01 Dev#= 12 Spd=12   MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs=  1
P:  Vendor=046d ProdID=c531 Rev=21.00
S:  Manufacturer=Logitech
S:  Product=USB Receiver
C:  #Ifs= 2 Cfg#= 1 Atr=a0 MxPwr=98mA
I:  If#= 0 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=02 Driver=usbhid
E:  Ad=81(I) Atr=03(Int.) MxPS=   8 Ivl=1ms
I:  If#= 1 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=00 Prot=00 Driver=usbhid
E:  Ad=82(I) Atr=03(Int.) MxPS=  20 Ivl=1ms

C2 (pc2)	51,5%
C3 (pc3)	0,7%
C6 (pc6)	0,1%
C7 (pc7)	0,0%
C8 (pc8)	0,1%
C9 (pc9)	0,0%
C10 (pc10)	46,7%

Baseline + Razer DAV2Pro mouse. Total power consumption: ~ 11.5 W

T:  Bus=01 Lev=01 Prnt=02 Port=06 Cnt=01 Dev#= 14 Spd=12   MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=1532 ProdID=007d Rev=01.00
S:  Manufacturer=Razer
S:  Product=Razer DeathAdder V2 Pro
S:  SerialNumber=000000000000
C:  #Ifs= 5 Cfg#= 1 Atr=a0 MxPwr=100mA
I:  If#= 0 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=02 Driver=usbhid
E:  Ad=81(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 1 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=01 Driver=usbhid
E:  Ad=82(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 2 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=01 Prot=01 Driver=usbhid
E:  Ad=83(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 3 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=00 Prot=01 Driver=usbhid
E:  Ad=84(I) Atr=03(Int.) MxPS=  64 Ivl=1ms
I:  If#= 4 Alt= 0 #EPs= 1 Cls=03(HID  ) Sub=00 Prot=01 Driver=usbhid
E:  Ad=85(I) Atr=03(Int.) MxPS=  64 Ivl=1ms

C2 (pc2)	51,5%
C3 (pc3)	0,9%
C6 (pc6)	0,2%
C7 (pc7)	0,0%
C8 (pc8)	0,1%
C9 (pc9)	0,0%
C10 (pc10)	46,3%

Related issue: High power consumption when using USB1.1 serial devices under Linux.

Any information on this and possible workarounds would be greatly appreciated.

Best regards,

Daniel

[KrissyG]

---

@Dan0987 wrote:

Hi @ACarmona_Intel,

---

• What prompted you to measure the power consumption?

---

Energy prices in Europe have recently soared, making the replacement of energy inefficient equipment a necessity. I measured all devices that could significantly affect my electricity bill. One of the main motivations for the computer upgrade was energy efficiency.

Idle power consumption of a PC without the monitor will be 20~70W, 20W if you set it on power saving and the PC does have 1 RAM stick and a very efficient GPU. 1 SSD or HDD and not too many fans either.
Measured at the electric socket with a fairly precise power meter. 

Assuming you let your monitor-less PC run 24/7, the average power consumption will be closer to 25W on idle.
The difference you seem to have measured was 6W?

The real difference is close to 25% on idle. It will be far less than 10% if you do anythign with your PC.

24/7 and 20W -> 0,48kWh, at a price of lets say 0,40Eur/kWh, that would be close to 0,2Eur per day.....and 73Eur worth of electricity in 1 year. (it will be much more than that)

Then the difference of 25% will be like 18Euros on a scale of 1 year - again, without touching the PC for 1 whole year.
That is extremely insignificant. 


Gaming PCs take easily 70~150W on idle, where for example XTU will show 5W CPU power usage. 

[Dan0987]

It is not insignificant when you take into account the scale of the problem (likely every system with 600 and 700 series chipsets is affected) and the fact that it is simply energy unnecessarily wasted.

[ACarmona_Intel]

Hello Dan0987,

Thank you so much for the response.

Again, we will do further research on the issue that we have and post the response on this thread once it is available.

Thank you, and have a great day!

Best regards, 

Carmona A.

Intel Customer Support Technician

[ACarmona_Intel]

Hello Dan0987,

Thank you so much for the response.

Again, we will do further research on the issue that we have and post the response on this thread once it is available.

Thank you, and have a great day!

Best regards,

Carmona A.

Intel Customer Support Technician

[ACarmona_Intel]

Hello Dan0987,

Again, thank you so much for patiently waiting for our response.

Our senior engineers have already conducted an intensive investigation into our case, and according to them, the problem is not related to the processor or the chipset. In regards to that, we recommend that you contact your USB manufacturers.

Thank you for your understanding, and have a great day ahead!

Best regards,

Carmona A.

Intel Customer Support Technician

[Dan0987]

Hi @ACarmona_Intel,

Thank you for your time.

It is extremely unlikely that devices from different reputable manufacturers, produced 10 years apart would have an identical flaw. I will purchase a PCIe USB controller and report the results as soon as possible.

Best regards,

Daniel

[ACarmona_Intel]

Hello Dan0987,

We sincerely appreciate your feedback.

We will patiently wait on the outcome of your new purchase PCIe USB controller, and if you have any further questions, please let us know. We will be happy to help you.

Thank you and have a great day ahead!

Best regards,

Carmona A.

Intel Customer Support Technician

[ACarmona_Intel]

Hello Dan0987,

We would like to check on the outcome of your new purchase of a PCIe USB controller? 

We look forward to your response!

Thank you, and have a great day!

Best regards,

Carmona A.

Intel Customer Support Technician

[Dan0987]

Hi @ACarmona_Intel,

My apologies for the delay. I haven't yet had free time to do some research on which chipset will work properly on Linux. Please bear with me, it may take a while.

Best regards,

Daniel

[ACarmona_Intel]

Hello Dan0987,

We understand. and thank you so much for the response.

No worries, We will patiently wait for your update.

Thank you, and have a great day ahead!

Best regards,

Carmona A.

Intel Customer Support Technician

[ACarmona_Intel]

Hello Dan0987,

Again, we are just checking in with you to see if you already have your new PCIe USB controller.

Thank you, and have a great day! 

Best regards,

Carmona A.

Intel Customer Support Technician

[Dan0987]

Hi @ACarmona_Intel,

I am sorry, this is going to take more time than I expected.  I will order something on Aliexpress (unfortunately the only place that offers any selection of cards on µPD720201) over the weekend, so I expect to receive it in late March or early April.

Best regards,

Daniel

[ACarmona_Intel]

Hello Dan0987,

Thank you so much for the update.

Will it be okay if you just create a new thread once you have received the PCIE controller? Once you have created a new case, please share this thread with us so that it can be used as a reference.

Thank you, and have a great day ahead!

Best regards,

Carmona A.

Intel Customer Support Technician

[ACarmona_Intel]

Hello Dan0987,

Again, we would like to ask if it will be okay if you just create a new thread once you have received the PCIE controller. Once you have created a new case, please share this thread with us so that it can be used as a reference.

Thank you, and have a great day!

Best regards,

Carmona A.

Intel Customer Support Technician

[Dan0987]

Hi @ACarmona_Intel,

Sorry for the late reply. It is not a problem for me at all, but it would probably be better for users coming here via Google to have everything in one place in case we find a solution or at least a workaround to the issue.

Best regards,

Daniel

[ACarmona_Intel]

Hello Dan0987,

No worries about the delayed response. I completely understand, and it's not an issue.

I agree that consolidating all relevant information in one location would be beneficial for users who find this thread through Google. This way, if we come across a solution or even a temporary fix, it will be easily accessible to everyone in need.

Thank you for your suggestion, and let's continue working towards resolving this matter. In case you're going to create a new thread, you can put the old link to the new thread so we can go back to the previous thread that you created.

Thank you, and have a great day ahead!

Best regards,

Carmona A.

Intel Customer Support Technician

[Dan0987]

Hi @ACarmona_Intel,

I did initial testing with the PCIe USB controler. Unfortunately, instead of the one I ordered, on a Nec/Renesas chipset, I recieved one on a VIA chipset, probably one of the worst available. The card works, but when installed the deepest state of the CPU package is C3. I have no idea whether this is the fault of the USB controller or the motherboard's buggy firmware. 

Anyway, with the card installed, at idle, with no USB devices connected, the system's power consumption is 12.6 W. When a keyboard (12 Mbps) is connected, the power consumption increases by only 200 mW, to 12.8 W. For comparison, under the same conditions, when connected to a USB port attached to the B660 controller, power consumption increases by 3.1 W, to 15.7 W. I observed almost identical results on Windows 11.

Best regards,

Daniel

[bjp83]

Hi all,

I confirm the problem with the following hardware:

- CPU: i7 14700

- motherboard: Gigabyte B760M DS3H AX

As USB device I tried various ways of plugin in a SONOFF USB zigbee dongle (which has a c210x USB-to-UART chip as entry-point):

- in a USB2.0 port of the motherboard

- in a USB3.0 port of the motherboard

- via a 4-port USB2.0 hub

in all cases, whenever I start using the USB device, the lowest package C-state goes from C10 to C2, and my idle power goes from 6-7W to 21-22W.

I can start using it by starting a Virtual Machine running Home Assistant OS (host OS is proxmox VE 8.2). When I stop that virtual machine, I get back C10. If I do the same without the USB stick plugged in, the VM runs, everthing works but my zigbee network, and C10 is still available, with low idle power (7W).

Workarounds I've seen on forums: use a Raspberry Pi to plug-in the SONOFF stick, and use ser2net to access it over the network. In that case, C10 is possible again apparently.

Which information/tests could I do to help Intel figure out the issue?

Best regards

[KrissyG]

---

@bjp83 wrote:

Which information/tests could I do to help Intel figure out the issue?

Best regards

---

is it intel tho? 
Why not Windows or the motherboard itself?

Here an example, same CPU, exactly same OS version and OS build, same hardware....except the motherboard and maybe some drivers, since different motherboard would have different drivers for same hardware.

Why do i get different values?

Gigabyte said this: 
"Dear customer,

Thank you for emailing GIGABYTE.Our team did answer like this:

There are many factors would cause TDP idle wattage such as CPU and memory.

Did customer try to use another CPU and memory model to check?

The Z790 chipset is better than B760. Hardware design need more power wattage to maintain system stability.

Please kindly enable C-states control and enable related bios settings such as CPU Enhanced Halt (C1E), C6/C7/C8/C10 State Support.

Please disable Intel(R) Turbo Boost Technology and Intel(R) Turbo Boost Max Technology 3.0.

Please also can try to lower CPU voltage to check.

Regards

GIGABYTE Team"

I literally proved the BIOS to be at fault, and they said it's hardware design.
Not to mention, the latest BIOS fails to allow XTU overclock all cores, one core is stuck at 5,3GHz or less than that.

[bjp83]

I'd say:

- not Windows, because I use proxmox (Linux/Debian based).

- interesting to mention I do have a very similar problem on a different board. Which to me points towards CPU/chipset combination.

If it's a software issue (which it could be in different ways: operating system, BIOS or CPU microcode), then it's a chain that fails. To me Intel and Gigabyte can point at each other, talk to each other, but in the end one/the two should come up with a solution.

If needed, OS support (Linux Kernel team) could be involved too.

I just hope a solution comes. It's a pitty to have idle power go from 6-7W to 21-22W because of this.

[Dan0987]

---

@bjp83 wrote:

Which information/tests could I do to help Intel figure out the issue?

---

I think it is obvious that Intel is not interested in solving this issue or even acknowledging it. At least until someone like Wolfgang brings it to the public's attention.

@KrissyG  wrote:

is it intel tho? 
Why not Windows or the motherboard itself?

---

The problem is reproducible on Linux and Windows, on boards from different manufacturers, on the initial and latest BIOSes. The only common factor, other than USB devices, is the 600 and 700 series chipsets. Your issue was unrelated.