Bios corrupted, CPU downclocking under load??

[R-T-B]

Lousy VRM "design" would not cause throttling to start now when it used to work fine.

If one of the VRMs are finally giving out it would.

This is an old board. It's not unreasonable to assume that.

 

[EarthDog]

Run intel xtu and get a load on the system. Xtu lists throttle reasons on the bottom and will light up yellow.

Mobo is weksos too...old... could be crapping out.

 

[Bill_Bright]

Or maybe i can get my friend's Evga 650 gold nova g2 which is same output as mine, but more efficient and newer, so i could do that.

At least for a test to see if the same problem occurs. In the meantime, what does HWiNFO64 (select sensors only) say for your voltages?

If one of the VRMs are finally giving out it would.

You took that out of context. Please note what I said about that in its entirety - that is, what you quoted and the part you left out. I said,

Lousy VRM "design" would not cause throttling to start now when it used to work fine. Lousy VRM "design" would be apparent the first time you tasked the system in this same manner.

Poor quality components used in the design could result in premature failure or degradation of parts.

 

[R-T-B]

You took that out of context. Please note what I said about that in its entirety - that is, what you quoted and the part you left out. I said,

I did, sorry. Didn't mean to, I think my crappy cellphone truncated your post.

 

[Bill_Bright]

No problem. Thanks for clarifying.

 

[Filip Georgievski]

Run intel xtu and get a load on the system. Xtu lists throttle reasons on the bottom and will light up yellow.

Mobo is weksos too...old... could be crapping out.

Here are results during test:

Here is after a few seconds:

And it goes on and on just looping like that, i can hear the fans ramping up and down and i know when it dropped.

 

[EarthDog]

I dont see the parts on the bottom for power limit and current limit...... did they turn yellow during testing?

That said, it doesnt look like its reading things off that board/cpu well either..

 

[Regeneration]

XTU says Turbo Boost is enabled, motherboards throttle CPU clock if it exceeds TDP/TDC limits. Set Ratio manually and disable Turbo boost, or use ThrottleStop (if you can't do it through the BIOS) to increase TDP/TDC if you wish to OC with Turbo on. Try to toggle Xtreme Phase Full Power Mode.

 

[Filip Georgievski]

Wow, i just realised that before reading post, but funny thing is SpeedStep is disabled, and with it Turbo as well.
So how does it try to turbo without SpeedStep, im kinda confused.

 

[Regeneration]

Could be a BIOS bug or XTU reporting error. Turbo needs SpeedStep and C-states to work correctly.

 

[Bill_Bright]

Also, your HWiNFO screen shots don't show your voltages. Scroll down to your motherboard section. They should be listed there.

According the ATX Form Factor standard, PSUs must maintain voltage tolerances within ±5% of required specifications.

Acceptable Tolerances:

12VDC ±5% = 11.4 to 12.6VDC
5VDC ±5% = 4.75 to 5.25VDC
3.3VDC ±5% = 3.14 to 3.47VDC​

 

[EarthDog]

Why would a psu cause a cpu to throttle? At worst it gets unstable before it just shuts off. Clocks wouldnt lower, would they???

 

[Bill_Bright]

It is just good to verify supplying good power to electronics when there is a problem that could be related to hardware.

 

[Regeneration]

Last time I had unexplained throttling like this, it was on Q6600 and some cheap MSI motherboard, due to a poorly coded BIOS. I solved it by cross-flashing the BIOS to a more updated one for a different motherboard, but with a similar design.

 

[Filip Georgievski]

Just finished a stock run on Intel XTU, max boost was 2.9Ghz and no downclocks, nothing, kept going the whole time with temp sitting around 50C at load.

I think i should reset the CMOS on this board and do a flash on older BIOS, and then back to the newest one, it is somehow buggy, using Turbo even when it is disabled.

 

[eidairaman1]

If you think the bios is corrupt.

Restore the bios to defaults, REFLASH IT then restore defaults again, make sure cpu cooler doesnt have plastic cover on it or the tim. Reapply tim, remount cooler correctly, ensure fans are plugged into correct headers, ensure good airflow and lower ambient temps.

Run rig, then run tests again in a week, recording your ambient temps and then system temps for today and then 7 days from now, continue to do that till you are satisfied.

 

[Solaris17]

Have you checked the power settings in windows?

 

[EarthDog]

Just finished a stock run on Intel XTU, max boost was 2.9Ghz and no downclocks, nothing, kept going the whole time with temp sitting around 50C at load.

I think i should reset the CMOS on this board and do a flash on older BIOS, and then back to the newest one, it is somehow buggy, using Turbo even when it is disabled.

Try it with your overclock and see what happens...

 

[Mussels]

this really sounds like VRM throttling to me, stop relying on software readings for them

wedge a fan in over the VRMs and run the tests again to see if it throttles

 

[basco]

maybe there is a cooler or thermal paste problem when your cpu starts from 50° idle

 

[EarthDog]

XTU didnt show it, neither does hwinfo. Where are you seeing its past tcase? Typically we see core/tmax temps, not tcase....

maybe there is a cooler or thermal paste problem when your cpu starts from 50° idle

that wasnt at idle... looking at minimum cpu loads.

this really sounds like VRM throttling to me, stop relying on software readings for them

wedge a fan in over the VRMs and run the tests again to see if it throttles

look at xtu... it says it there if he scrolls down in the monitoring part of xtu. Hwinfo shows none as well. Though, i suppose it doesnt hurt to try that either.

 

[Reeves81x]

XTU didnt show it, neither does hwinfo. Where are you seeing its past tcase? Typically we see core/tmax temps, not tcase....

that wasnt at idle... looking at minimum cpu loads.

look at xtu... it says it there if he scrolls down in the monitoring part of xtu. Hwinfo shows none as well. Though, i suppose it doesnt hurt to try that either.

73C in his screenshot. Google it. depending on the mobo it will throttle at tcase, which is 72.7C which is exactly what i am seeing here. It's supposed to be to keep from damaging your computer.

"Tjunction Max (TjMax) is different than the TCC Activation Temperature. TCC offset represents where the processor starts to throttle, relative to the TjMax value. Tjunction can be the same temperature as TCC if the TCC offset equal to zero."
https://www.intel.com/content/www/us/en/support/articles/000005597/processors.html
corroborating info:
http://www.overclockers.com/forums/...Buying-Memory-for-Old-System-(DDR3-P55)/page2

 

[EarthDog]

I know exactly what it is and means. I was unaware that value was tcase in hwinfo.

Nothing is catching it though is my point. Hwinfo and xtu doesnt show any thermal throttling which is what that will cause.

Edit: and when i just googled what cpu max was, the author of hwinfo said its the hottest temperatures of all cores, which, as you know, would not be tcase as it is usually lower than tjunction which is what most applications seem to use for temps. So it seeks that isnt tcase and it isnt throttling.
https://www.hwinfo.com/forum/Thread-What-CPU-temperature-reading-should-I-use

 

[Reeves81x]

I know exactly what it is and means. I was unaware that value was tcase in hwinfo.

Nothing is catching it though is my point. Hwinfo and xtu doesnt show any thermal throttling which is what that will cause.

i think you misunderstood. some Mobo's set an TCC offset that brings the throttle temp to 73c tj or thereabouts, Dont ask me why.. its just a stupid thing that they did. Basically it throttles at 73 tj and will actually shut down the pc at 76-78 tj in this configuration. Just cool it down, keep it below 72 and tell us if that works. it Could still be hitting the wall for amperage draw in the bios due to the crappy vrm, would be nice to see some figures on how much its drawing. But i have now read 4 threads with the same gen cpus having throttling at 72-75c
also read a case of c3 a c6 states causing this.

https://forums.overclockers.com.au/threads/i7-860-throttling-down-at-70-73-degrees.848727/
http://www.overclockers.com/forums/...Buying-Memory-for-Old-System-(DDR3-P55)/page2
http://www.tomshardware.com/answers/id-1893896/cpu-throttling.html
https://www.dell.com/community/General/Dell-7567-Throttling-73-Degrees/m-p/5187746
http://www.overclock.net/forum/5-in...28v-already-throttling-70-degrees-solved.html

 

[EarthDog]

Perhaps i did...? If that is true, why didnt xtu or hwinfo catch it though?

See edit above btw...