Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Vcore Down
#1
My vcore is low as seen in the picture, it's a bug?


Attached Files Thumbnail(s)
   
Reply
#2
That happens on CPUs featuring Fully Integrated Voltage Regulator (FIVR) when idle.
CPU-Z is probably filtering such values, or shows VID instead.
Reply
#3
aida64 too read fine the vcore, only hwinfo64 have a bad reading
Reply
#4
the problem is only in ite it8620e, watch screenshot
Reply
#5
AIDA64 reports VID instead of Vcore there. You can see this value in HWiNFO sensors as well.
Reply
#6
Hi Martin,

Firstly many thanks for creating & providing HWiNFO Cool , I find the program invaluable Wink .

Currently using v5.10-2700 I get vcore as 0.000v across all 4 cores at idle, in current & min column.

I have an i5 4690K + Asus Maximus VII Ranger (FW 3003), C-States/EIST are all enabled and I use adaptive voltage mode. I know CPU has a FIVR but with versions v5.06-2640 and lower, vcore is displayed and logged fine at idle.

Is there an explanation you have for this occurrence?

Also any bugs in v5.06-2640 which would warrant using newer versions over it?

Cheers.
Reply
#7
Is the problem occurring only at idle and at load you're getting correct Vcore values ? What values are you getting with older versions in idle? Can you post a few screenshots ?
Some mainboards report VID instead of Vcore when idle (values too low) because of this.
Reply
#8
Many thanks for a swift reply Smile .

Yes problem is only at idle, its not a biggie but as I use the program to log average as well, so that figure then is inaccurate on newer versions.

[Image: etCfc6c.jpg]

Orange box at top is VID (I have not renamed sensor, this is as HWiNFO detect and apply naming).

Red box is Vcore (I have not renamed sensor, this is as HWiNFO detect and apply naming).

I set a VID of 1.255V in UEFI, so VID up top is correct. The slight overvoltage in Vcore is due to adaptive voltage mode/AVX code stress program being used, so not an issue.

Only other change on my system between that screenshot and current system configuration is motherboard firmware, currently 3003, that screenshot is 2801.

I've had this motherboard since about March 15 and been using your program since then for many hours with differing firmware 2601, 2702, 2801 and had no issue on idle vcore. I don't recall an issue with reported idle vcore with program/firmware 2902 but will check.
Reply
#9
Thanks for the screenshot, can you please make a similar using the latest HWiNFO Beta ? I want to see the difference between those versions, as it's a bit odd that they behave different.
Also, do you maybe see a difference in CPU load when idle and using v5.06 and later ones ? My theory is that v5.06 might cause a slightly higher CPU load by itself which prevents the CPU entering low-power states (with Vcore being very low), while later versions have a very low CPU overhead and thus the CPU enters low power states and Vcore is down.
Reply
#10
Will try beta ASAP and report back, will check CPU load when idle with HWiNFO running.

Your support and swift response are phenomenal! Cool
Reply
#11
Hi Martin,

Got to the bottom of what I thought was issue.

Originally HWiNFO did not show VCORE at idle ie 0.xxxV (~Q1 15). I thought it was issue and was new to this platform, later it was showing a VCORE reading similar to VID at idle. I thought a bug fix had occurred in bios / HWiNFO, then after a reinstall of Win 7 x64 it stopped working (~Q4 15).

I have not been able to make older editions of HWiNFO show VCORE at idle on new install. I think between Q1 15 and Q4 15 my install of Win x64 had a service / app creating loading which then made VCORE similar to VID at idle.

Cache voltage also now is at 0.xxxV when at idle, this is correct operation?

Cheers for your support!
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)