Well I tried what @JLK03F150 found, but did not solve the issue. I opened a ticket at EA forums tho so maybe get something from them. I would like to ask if it is normal that my GPU "3D" usage is jumping to 100% for minor stuff like opening windows snapp tool to grabb a screenshot of browser with only one tab opened. Sometimes even doing nothing on computer will still show some 15% to 30% "3D" usage sometimes for minutes. Is this normal? wouldn't this cause apex to crash or system reboot if "3D" usage is being used so hard for minor stuff? is this normal or not?
PD: Thank you all for your help, I know this forum is not specific for this kind of issues but at least I get knowledge from here when asking stuff !
Hi guys, I was testing some other games and got blue screen on PUBG, I know it is not the "buzzing sound with black screen and then reboot" that I got in Apex legends but I thought it was only Apex having issues, so there is something with my system I got to find. So far I have replaced PSU, SSD. Did a memtest86 and when I came back to check "it was taking some time 1 hour and 30 min" it was on the last pass and under "error" it got 0. I went to bathroom and after coming back computer was in windows... it was normal if the test is done that it will restart automatically? or did the test fail and that's why I found my pc in windows instead of memtest86?. The system reboot or blue screen it's only happening in demanding games like Apex or PUBG. Minor games will cause no reboot or BSOD. Using pc for watching movies or surfing the web works pretty good. Games run fine tho in terms of FPS, performance is awesome but worthless with these crashes. Could you guys tell me anything I can try in bios? I read some people had this issue with Ryzen 7 5800X. I'm running out of ideas. I was even considering giving a try to "pc doctor" it charges you for 30 USD or something but reviews say it may find issues pretty good. The thing is, before upgrading my PC I was playing Apex with no system crashes with the same GPU I have now, so I don't believe it could be the issue but IDK. The rest of the hardware is all new.
Is this the right windows? I just went into BIOS and set curve optimizer to disabled, somehow it was set to -15 all cores. It was supposed to undervolt this way and help on temp. I also disabled Global c state and for some reason temps are a bit lower than before changing these settings. I still see high voltage tho like close to 1.5v but I don't care at this point I want system stability =((( I'll report if it keeps crashing or not. Also do you people download chipset driver from AMD or Motherboard website? AMD comes with Ryzen Balanced Power Plan and Asus's one didn't box that one.
I got some WHEA Logger tho... these are CPU specific I believe...... so it may be my Ryzen responsible for all the crashes....
And with Curve Optimizer disabled and global c state disabled I get these numbers with AIDA64 stability test, 4,5ghz seems low? or is it normal? Well I'll test if it still crashes or not and report back.
The point is that it uses quite high voltage while doing not much, like moving around windows or in game lobby "1.45v 1.48v" but quite low under load "1.38v" Will I need to use manual voltage control and use LLC to get lower voltage on idle "low use" and higher voltage under load?
PD: I always wanted to know, the "A" in picture below is for amperage? if so it has something to do with PSU amperage? mine has this labbel, 54A alone is far less than what I see in HWmonitor, or it is combined with the dual 20A?
PD: Something very strange, the fact that after disabling Curve Optimizer I get higher voltage but quite lower temps!? how is that possible? in fact prime95 doesn't push temps higher than 65ºC while with Curve Optimizer and Global c state prime was always pushing 81ºC to 85ºC. Even gaming is hiting lower temps, what's going on? So far no crashes as well, I'll keep testing...
Is it true that voltage above 1.35 will cause degradation of the chip? I read many post pointing this out.
Not like that. A friend told me to test everything on auto. I tested and I still got reboot in apex so in the Middle at some point I re enabled curve optimizar negative values as ^it was still crashing^ for lower temps. But that's why I am still trying stuff. Yesterday played apex for like 2 hours with no crashea But I need to test harder.
What was the crash code for the app when it happened? I'm not entirely certain what I'm seeing above and if it relates to that time and what happened... sorry.
It's so random, this problem... lol. But doesn't crash in stress tests.........lol
Did you try with the memory at auto too? How about with one stick (in the proper slot for one stick, see manual).
What was the crash code for the app when it happened? I'm not entirely certain what I'm seeing above and if it relates to that time and what happened... sorry.
It's so random, this problem... lol. But doesn't crash in stress tests.........lol
Did you try with the memory at auto too? How about with one stick (in the proper slot for one stick, see manual).
I did the mem86 test and it looked fine. I don't think it's ram related. Just theory but with global C state enabled and curve optimizer -15 all cores somehow the mechanics were generating idle stages of 0.2v and ramp from there when 4,85ghz was required. This may have been generating issues, because somehow it may just draw more power or more voltage in these conditions. While setting curve optimizer to disabled, and disabling global c state, voltage won't go lower than 0,9v and ramping is now closer to target vcore, it may explain why do I have lower temps now. Basically trying to undervolt just by curve optimizer was actually outvolting in ramp situations. "all theory, but so far no crashes" I'll try to play a lot as soon as I get time because it's annoying to be concerned about these reboots with new hardware.
For instance, running prime95 now with very cool temps "no more than 65ºC" I get this values using hwinfo as max values for what this software reports about voltage. If you see, it gets lower 1.431 V compared to curve optimizer -15 all cores with global c state enabled, which was giving me 1.487 very often. I'll keep doing test, but so far even if it crash again after deep testing; there is an impact this values got because I'm in a new record with no reboots.
Today I got a reboot while Playing Apex, it was like 3 days crash free but it did happen again. So I think it is cpu related. I don't know what else to try, should I disable CPB to run at 3,6ghz and check if it will not crash anymore? or should I try to give it more voltage?
This latest reboot gets this error:
I have everyting default in bios like in this picture:
As you can see on bios a lot of CPU voltage is noted, and temps while on bios are 50ºC
Ironically I never see this 1.48 while running any stability or stress test on the CPU. Should I just RMA my CPU? I hope not, IDK how would I do that I bought it imported it so I didn't bought it in my country.
I don't know what happens, but now I was testing with prime95 and temps gets to 64ºC for like 10 min, and then in one second they ramp to 90ºC .....
I also see 140A and in ryzen master I believe I saw some 140A limit in red
before trying to RMA this CPU can someone tell me how to set manual clock of 4.6ghz ? I would like to try static 4,6ghz with some acceptable voltage and see if it solves the issue, if not RMA. First AMD cpu in my life to give me problems.
I don't get it, if everything is on auto how am I overclocking? I thought running top clock at 4,6 is down clocking as it gets to 4,85 while gaming but I see 4.5 4.6 on stress test. So maybe this cpu silicon is garbage and cant do 4,85 that's why I was asking. All hardware pieces are new. When I bought my fx-8350 things were just simplier.
I would raise the stock voltage before I would mess with anything else. We 33 posts in and I don't recall if you tried a positive offset or setting it static and increasing, but... I really don't think it's the CPU regardless.
I would raise the stock voltage before I would mess with anything else. We 33 posts in and I don't recall if you tried a positive offset or setting it static and increasing, but... I really don't think it's the CPU regardless.
Why do you believe it is not cpu related? I'll let run OCCT for 1 hour. so far this is the values from it, are these similar to yours or anyone u know that got a 5800x?
I'm searching internet on how to do what you suggest, I don't know how many options I have to disable in bios in order to be able to use positive offset, it can be used just like that? or I need to disable these boosting crap PBO CO so many options god.
I tried prime95, but as I said after a few minutes it will sky rocket my temps to 90ºC no bueno. It seems OCCT don't ramp temps like that so I'm trying to first check that CPU is not defective or faulty, because RMA would be hard in this particular case. So I have to somehow find a way to prove it's faulty, and I don't come with more ideas than running stablity test.
What test in P95 are you running? Blend is what you should be running (tests, CPU/RAM). If you're just looking at the CPU, run Small FFT. If it reboots there, it's power or voltage or something. I haven't used OCCT in......a decade.... so I have no idea how good it is. At this point, For my uses, AIDA64 and it's stress test(s) does well for me.
Worth noting, 90C is fine for that CPU. It starts to throttle at 100C.
I just don't think your CPU is faulty. Please test to confirm/deny, but... I don't think it is. I'd lean more towards the mobo acting wonky first.
Did you try with another PSU? I know that's a high quality SFlower unit, but... it's worth a try.
As far as the offset, It should be in the tweaker menu where you set the voltage. It will have options like offset/adaptive/override etc.
What test in P95 are you running? Blend is what you should be running (tests, CPU/RAM). If you're just looking at the CPU, run Small FFT. If it reboots there, it's power or voltage or something. I haven't used OCCT in......a decade.... so I have no idea how good it is. At this point, For my uses, AIDA64 and it's stress test(s) does well for me.
Worth noting, 90C is fine for that CPU. It starts to throttle at 100C.
I just don't think your CPU is faulty. Please test to confirm/deny, but... I don't think it is. I'd lean more towards the mobo acting wonky first.
Did you try with another PSU? I know that's a high quality SFlower unit, but... it's worth a try.
As far as the offset, It should be in the tweaker menu where you set the voltage. It will have options like offset/adaptive/override etc.
Thank you for that info, very useful. The 1 hour OCCT test finished with no issues. I aready replaced my PSU, older was an EVGA 600W that I got recommended by a someone in this forum for my old fx rig. I bought this SFlower because friends of mine got that one and they are getting no issues on high end hardware, like rtx3090 etc. I was worried about 90ºC because I thought it was possible to avoid that by air cooling but I think it's not the case with these cpus. I think I'll just have to ignore and assume that 90ºC ain't that bad. Testing with prime95 "I was using blend default mode" It already passed the CoreCycler-v0.8.2.4 core by cores test tho. Because if it is not generating any errors during all these cpu stress test, and crash only happens in games, either could be other issue. So I also disabled xmp for RAMS, and let them just default, they are running at 3.200mhz just to make sure, and VGA shouldn't be a problem since never crashed anything with the old rig, plus I don't have any older VGA available at the moment for testing. So I'll report back, and if it crash, I'll bump voltage with offset.
That screams the game/drivers with those games/GPU, etc. I mean, it's passing all the hardware stress tests....at least CPU and RAM...and in those you aren't using the GPU. You use the GPU and crash... right? So it's something along those lines, I'd image. Check to see if your GPU has a BIOS update (doubtful). If it does, may want to do that too.
But yeah, sort of hard to follow here... If it was me, I'd just remove this drive, put windows on a fresh drive, install the drivers... install the game(s) and try it again. Somethng tells me it's a click fest of software so who knows what's going on by now with that windows install. But the good news is, doing that leaves your current install intact just in case that isn't the issue.
That screams the game/drivers with those games/GPU, etc. I mean, it's passing all the hardware stress tests....at least CPU and RAM...and in those you aren't using the GPU. You use the GPU and crash... right? So it's something along those lines, I'd image. Check to see if your GPU has a BIOS update (doubtful). If it does, may want to do that too.
But yeah, sort of hard to follow here... If it was me, I'd just remove this drive, put windows on a fresh drive, install the drivers... install the game(s) and try it again. Somethng tells me it's a click fest of software so who knows what's going on by now with that windows install. But the good news is, doing that leaves your current install intact just in case that isn't the issue.
Yes, it may look like that. But I went around CPU because of the curve optimizer behavior. The main issue here is a reboot while playing apex, so when I set curve optimizer to -15 and play Apex, it will crash exactly the same way giving the same WHEA error for specific core, which in EVENT VIEWER is described as ID 10 but I don't know how to convert this ID to the actual core. So the error it's the same. The concept is that when I set -15 on CO and play Apex, it will crash the same way, but just after 10 to 15 min playing, it will reboot as soon as that, sometimes even sooner. So if I revert everyhing back to AUTO it will crash every couple of days, but the error it's the same. That lead to me to believe it is CPU related, but it could be as you said many other thing that make it look like it is CPU related. So after having nothing to try I went into bios press F5 loaded optimized default and then set CPU VCORE to 1.297 and the multiplier to 46.7 so now I was testing what has changed? well voltage looks obviously lower, but something changed; if I run cinebench for instance, it will pass 90ºC easily, I stoped test when I saw it about to hit 97ºC. With stock settings voltage is higher by a lot, even under load; but temps on cinebench were literally locked to 90ºC it would drop everything but it won't pass 90ºC. As soon as I set manual voltage and manual multiplier for CPU, this is no longer valid, and scares me because I don't know if it will stop at some temperature but 97ºC seems like wrong. In the other hand I see static 4,67ghz on everymonitor, there is no more idle clock which I would miss tho, however in hwinfo the effective clock speed always seems to go close to 0 when idle, and even on stability and stress test I don't see effective clock being that much close to 4,6ghz. This comes so far with impressive stability, I'm testing Apex and no crashes so far but I'll have to keep this settings for a while until it crashes. If it doesn't crash, it could be a possible valid configuration, so if that were the case; I'm forced to live with the stuck static 4,6ghz or is there any workaround to have the best of bost worlds. Maybe this chip just need to run lower than 4,85 "Will check if Apex doesn't crash anymore"
PD: Here, you can see effective clock quite low when playing apex for 2 hours.
I didn't set any custom soc voltage, is that okay? I just set vcore in bios and cpu multiplier...
Is Ryzen Master showing this CPU voltage as "CPU Core VID (Effective) from hwinfo as shown below???:
hwinfo:
If this setup proves to be stable for Apex that would be dope, but still bothers me if effective clock is low how much performance I'd be losing. In Apex I can't tell the difference, I would say it runs a bit faster but may be visual missread. "I only use ingame FPS" so I'm kinda used to what FPS do I get in most common parts of the game, and I see no performance drop.
PD: In fact I was running AIDA64 benchmark and pc crashed with CPU TEMPERATURE ERROR "THERMAL SHUTDOWN", which never happened with stock values but higher voltage. So how do I keep the 90ºC thermal limit that I lost after setting manual vcore and multiplier for CPU?
Your PC would be a slug if it was running that slow. If you're getting the expected performance, that's clearly wrong. Don't worry about effective clock. all these values are making your head spin, lol. You should only be running one monitoring software at a time where possible. Ryzen Master is typically the most accurate compared to the others.
I believe Ryzen master and VID are one in the same. yes.
But let's think here... it's only happening in Apex, can we just consider Apex is borked until you reinstall windows on a different drive (and set the bios to optimized defaults).
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
!
