Asrock, wtf did you do the Z170 Extreme4+ bios??
Printed From: ASRock.com
Category: Technical Support
Forum Name: Intel Motherboards
Forum Description: Question about ASRock Intel Motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=3604
Printed Date: 20 Jul 2025 at 11:13pm Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: Asrock, wtf did you do the Z170 Extreme4+ bios??
Posted By: Imakuni
Subject: Asrock, wtf did you do the Z170 Extreme4+ bios??
Date Posted: 17 Oct 2016 at 4:24am
Okay, so yesterday I saw ver. 7.0 was out and decided "what the hey, y not?". Then I updated, set back my previous OC and let my go back to 24/7 crunching Prime95.
Then today, just becuase it was a little hotter than usual, I decided to take a quick peek at my temps... and behold!, 98C across all cores! Voltage was being shot up to 1.360 average rather than the 1.296 max it was supposed to run at after LLC kicks in.
So yaaay, my PC's been running for almost a day pinned at 98C for whatever reason. 99% sure I've LLC back to the level I want it to rather than leaving at auto, but even if the setting did somehow reset to auto, it would cause vdroop, not overshoot (well, at least that's what happened previously anyway). Did Asrock change the LLC policies somewhere?
|
Replies:
Posted By: parsec
Date Posted: 17 Oct 2016 at 11:44am
Imakuni wrote:
Okay, so yesterday I saw ver. 7.0 was out and decided "what the hey, y not?". Then I updated, set back my previous OC and let my go back to 24/7 crunching Prime95.
Then today, just becuase it was a little hotter than usual, I decided to take a quick peek at my temps... and behold!, 98C across all cores! Voltage was being shot up to 1.360 average rather than the 1.296 max it was supposed to run at after LLC kicks in.
So yaaay, my PC's been running for almost a day pinned at 98C for whatever reason. 99% sure I've LLC back to the level I want it to rather than leaving at auto, but even if the setting did somehow reset to auto, it would cause vdroop, not overshoot (well, at least that's what happened previously anyway). Did Asrock change the LLC policies somewhere?
|
The first 7.xx UEFI versions for ASRock Z170 boards, are the CPU microcode updates for the new Kaby Lake processors.
Your experience with this UEFI version is different than mine, on my Z170 Extreme7+ board. I thought my CPU/core temperatures were lower than with earlier UEFI versions, with the same OC.
What UEFI version were you using before changing to 7.00?
What was your CPU temperature before you changed to 7.00, while running Prime95?
You told us nothing specific about your CPU voltage setting, but it sounds like it is set to the default, Auto, or Adaptive. You claim that is the same as you used previously?
Prime95 + Adaptive voltage + OC = extra high VCore when AVX2 instructions are being executed. LLC does not cause voltage overshoot, as you said, but Adaptive voltage will.
Since Haswell processors became available, Prime95 has fallen out of favor for stress testing by many OC enthusiasts, due to its excessive, unrealistic focus on AVX2 instructions. CPU temperatures reaching throttling levels became common. I never had temperature induced throttling until I used the updated Prime95 with AVX2 instructions on a Haswell CPU. You were 2C short of throttling.
It's all to easy to forget to reset something after a UEFI update, since any profiles you saved are gone after the update. Such as, your CPU cooler's fan speed configuration, although at 98C, they should be running at full speed. But your case fans are not automatically configured in the same way.
No details given about your PC, OC, so zero to go on. Change in LLC policies? No. We don't even know what you are using to monitor VCore. You could be seeing the max VID and not the true VCore.
------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: Imakuni
Date Posted: 18 Oct 2016 at 4:11am
" rel="nofollow -
parsec wrote:
What UEFI version were you using before changing to 7.00? |
2.80 and most versions before that.
parsec wrote:
What was your CPU temperature before you changed to 7.00, while running Prime95?
|
Running at around low 80s, peaking at 84~85. Not sure why that matters, though... problem here isn't the temperature by itself, but rather the excessive voltage that I know my cooler just can't handle VS the voltage I want for the heat I know my cooler can keep up with.
parsec wrote:
We don't even know what you are using to monitor VCore. You could be seeing the max VID and not the true VCore. |
HWMonitor. I'm not looking at VID. Mine always reports a VID of 1.170, regardless of load, bios setting... whatever is the case, it's always reporting 1.170. At any rate, I am going off of VCORE.
parsec wrote:
Prime95 + Adaptive voltage + OC = extra high VCore when AVX2 instructions are being executed. LLC does not cause voltage overshoot, as you said, but Adaptive voltage will.
Since Haswell processors became available, Prime95 has fallen out of favor for stress testing by many OC enthusiasts, due to its excessive, unrealistic focus on AVX2 instructions. CPU temperatures reaching throttling levels became common. I never had temperature induced throttling until I used the updated Prime95 with AVX2 instructions on a Haswell CPU. You were 2C short of throttling.
|
Don't worry, I know how P95 works. And I know it's a bad idea to let voltages on auto with my PC crunch it 24/7 for weeks on end without looking at temps. In fact, that's why I even messed with LLC to begin with, so that by putting a fixed setting, I would always have control of what's going on.
parsec wrote:
You told us nothing specific about your CPU voltage setting, but it sounds like it is set to the default, Auto, or Adaptive. You claim that is the same as you used previously?
LLC does not cause voltage overshoot, as you said, but Adaptive voltage will.
It's all to easy to forget to reset something after a UEFI update, since any profiles you saved are gone after the update. Such as, your CPU cooler's fan speed configuration, although at 98C, they should be running at full speed. But your case fans are not automatically configured in the same way.
No details given about your PC, OC, so zero to go on. Change in LLC policies? No. We don't even know what you are using to monitor VCore. You could be seeing the max VID and not the true VCore. |
1.275v set in fixed mode. LLC lvl 3 for a slight overshoot of 1.296v under load (VS 1.280 when load's not applied).
Oh yeah, and about that "LLC doesn't overshoot things"... well, it actually does. It might not do it for you, but I can assure you that, at the very least on my pc, it does. I don't remember the exact numbers (it's been a long while since I've tested this), but Lvl4 would cause vdroop, Lvl3 would be a slight overshoot, Lvl2 would overshoot higher and Lvl1 would just go ballz to the wallz on the overshoot. That's what happens with Vcore and both temperatures and power consumption readings from the wall were reflecting that.
In fact, once I've pannicked seeing 98C, I immediately fired the A-tuning utility and pulled down the slidder down to 1.23v. And sure enough, now it's happily running at 1.296v, temps back to what they were supposed to. Given that I can assure you it's not running in adaptive mode (even the A-tuning utility itself is reporting Fixed mode), I can only think of one other setting that could be causing this voltage spike: the LLC. Unless there's something else that messes with VCORE that I'm not aware of.
At any rate, I'm sure I didn't forget anything. Everytime I update the Bios, I make sure to go over every setting in every sub-menu to guarantee that I've set everything back, and this time wasn't different.
|
|