[SacraficeMyGoat]

Hey guys, had a strange issue today and was hoping to get some help.

My system has been running great since I built it a few days ago, but today it wouldn't turn on. There were no lights on the MB, like it wasn't getting any power whatsoever.

I ended up pulling the CMOS and reseating it and it powered right on. I'm confused why this happened.

Does anyone know a fix for this issue? I'd very much like to stop this from happening again in the future since I have to remove the GPU to remove the CMOS battery...

 

[SacraficeMyGoat]

Did you resolve it ?
Sounds like either you had a power outage - or someone unplugged your PC / shut down your PSU in between back then and today

It's working fine now, but there were no outages, someone messing with my stuff, or anything of the sort. It simply wouldn't show signs of power when I tried to turn it on this morning after shutting it down last night. Removing the CMOS fixed it, but hopefully it'll not happen again.

I also enabled ErP settings in the BIOS, as someone suggested this may help. It basically allows the system to FULLY shutdown and not use any power when it's off (thus not draining the CMOS.) We'll see with time if it actually helps.

I have run into another small issue though, my memory's RGB doesn't seem to turn on until I manually open RGB Fusion now. Wasn't doing this before. Not sure if there is a way to enable RGB Fusion to run on start up or something? May try a complete reinstall of it soon.

Edit: I'll need to do further testing, as the RGB has decided to turn on now on start up... >.<

 

[SacraficeMyGoat]

Someone send helppppp. Pulling my hair out over this...

I keep getting hard crashes while playing Destiny 2 (extremely randomly, I went like 2 days without an crash and today it's happened about 3 times.) I've tried multiple stress tests and can not seem to replicate the hard crash.... I was only able to replicate the crash one time, and that was with Cinebench R20. I've run AIDA64, Furmark, etc, etc. I just don't understand what is going on.. It always crashes when i'm AFK so I don't know if it freezes first then crashes, but I have caught it frozen one time and had to hard reset it. Temps are definitely not a concern, they stay around 70C. Aside from XMP enabled, everything else is stock I believe (radeon software is set to automatic/default, not sure if that actually overclocks the GPU or not.)

This is the error I get in Event Viewer after the crash (although the APIC ID # changes.) -

A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 14

I've yet to have it happen in other games, but I play Destiny 2 the most. I've sort of convinced myself it may be my memory kit (with the help of some redditors) when it's running at XMP, but honestly since I can't replicate the crash, I can't be certain. It did run for a few hours with XMP off w/o crashing, but again, it happens randomly so this may have just been a fluke. My memory kit is BL2K16G36C16U4BL which isn't on the QVL list (not many sticks on the QVL list tbh... kind of frustrating.) Came across a reddit thread and there are a few other people recently that have had this same issue (not destiny 2, but in general.) This leads me to believe it's not strictly related to Destiny 2. I have a replacement CPU coming Wed, and I'm thinking about ordering some memory (G Skill Trident Z Neo 32gb 2x16 3600 CL16) tonight so it will arrive at the same time and I can test both. If anyone is running that kit successfully, please let me know as it isn't on the QVL either!

I'd really appreciate some more input on this issue as I just built this PC from brand new parts about a week ago and it's extremely frustrating. Wouldn't be so bad if I could actually replicate the error and pinpoint the cause, but it's not always that easy is it...

Let me know if you need any more info, hwinfo logs, etc.

 

[SacraficeMyGoat]

You have to set Normal vCore with no offset, PBO Advanced with maximum configurable values for PPT/TDC/EDC, no scalar, no boost clock.
Then run P95 and check the voltages.



So it's 2 types of crash; one playing Destiny2 and the other when you are AFK? Or do you mean happens when you are AFK on Destiny2?

Seems to me a problem with VDDP/VDDG, post a Ryzen Master screenshot.

Sorry, no the crash usually happens when i'm AFK in Destiny 2. Not two seperate crashes.

 

[SacraficeMyGoat]

There are two versions of the trident z neo 3600. The one that ends with the GTNC is Hynix - I have those and run them at 3800 without issues. Destiny 2 plays perfect. There have been some reports of radeon GPU drivers causing drivers so hard to tell if the graphics card is causing you issues as well... I'm on a 2070 super.

Your RAM should be okay though... just try manually setting the Vdimm to 1.36V with your current settings and see how that goes?

Thanks, I have some 3600 Gskill Trident Z Neo (GTNC) coming tomorrow just incase it is memory related. I also set memory to XMP and manually set the voltage to 1.35V.

 

[SacraficeMyGoat]

Try to set VDDP 900 and VDDG 950.
Both in CBS > XFR and AMD Overclocking.

Could be the drivers but I'm not confident, usually it doesn't relate specifically to low load.

I will try this, but could give tell me how to change these settings please. I'm not familiar with AMD software yet. Do I need to do this in BIOS or Ryzen Master?

 

[SacraficeMyGoat]

Upon Googling your specific Event Viewer Codes I came across this site:https://social.technet.microsoft.co...vent-id-18-microsoft-windows-whea-logger.aspx

doesn't look good. I see you already have replacement CPU coming so hopefully that will resolve your issue.

As for RAM passing mem stability tests isn't enough as 1 borderline iffy setting will be enough to crash pc while gaming. You should add a few AAA games to your stability testing methodology to ensure RAM can handle heavy workloads. This is important because RAM issues on Ryzen often present while gaming causing users to look at other things besides the RAM as the cause.

Before I built my 1st Ryzen build back in 2017 I used to think that if I was able to post, install Windows, and boot into Windows everything was all good. Boy, I couldn't have been more wrong, lol.

There was an issue with 1st gen Ryzen freezing at idle because vcore would drop down below .0400 and some PSUs didn't like that at all. As a fix they added a PSU Idle power setting in bios that lets you choose "typical" and "Low". The "Typical" setting keeps vcore from dropping below .800 which keeps some older PSUs happy. I'm sure this isn't your issue as it only occurs for you while gaming and would hard lock the PC while idle at desktop needing a cold reset.

Any stress test programs I can use that will put the same amount of stress as a AAA game title? Going to play a few games today and see how things go, but when i'm not playing i'd like to put a synthetic load on.

 

[SacraficeMyGoat]

From the specifications page

Master v1.1



Master v1.0

Interesting. How can I tell if I have v1.0 or v1.1?

 

[SacraficeMyGoat]

Board revision is printed on the very bottom left corner. I doubt there was much changes in that area of the board anyway, probably just different testing done on 1.1, my 1.1 board seems more fussy about memory then my 1.0 board.

Thanks. Mine is 1.1 as well.

 

[SacraficeMyGoat]

Hey guys, few questions for you.

1)Wondering if I should update to F20a from F11 (i'll just be using XMP @ 3600 and maybe enable PBO, nothing else.) I've been having random crashes that I haven't been able to sort (replacement CPU and new set of memory came in today, so i'm going to swap the CPU out.) Will hold off on replacing the memory for now as I have no idea what is causing the crash since I can't replicate it and they're so random.

2) How can I download F11 BIOS on my 2nd BIOS chip and the new BIOS on my 1st chip? That way if there are any issues, I can just go back to F11.

3) What's the best method for upgrading the BIOS (just download the file and run?) Any steps/precautions I need to take?

4) I'm currently using the Gigabyte chipset drivers, is it better to use the AMD drivers? If so, can I just download them over the current chipset drivers?

5) I have v1.1 of the motherboard, do I need to download the drivers/BIOS from the v1.1 page or are they the same as v1.0? I hadn't realized this when I first built my system and downloaded the v1.0 BIOS/drivers, so I don't think it's an issue, but want to make sure.


Sorry, I know I have TONS of questions, but i'm just trying to make sure I do everything right..

 

[SacraficeMyGoat]

If this is the master, the easiest way to control BIOS flashing is just put the board into single BIOS mode and use the dip switch to select which is active/flash to whatever version you want to main/backup.

Not really a best method, some people swear by using QFlash+ (Flashback on ASUS, etc.) but I have had issues getting it to work on this board. I just end up going into the BIOS itself and using QFlash from there, easy enough.

As for drivers Gigabyte like most vendors seems to post whatever drivers they tested at the time, and generally do not keep those up to date, best to grab chipset drivers directly from AMD: https://www.amd.com/en/support/chipsets/amd-socket-am4/x570. Same for bluetooth/network/etc. from their respective vendors.

Board revision doesn't matter, since the hardware is the same from a driver perspective.

I'm currently on single BIOS mode (haven't messed with the switches since receiving the board.) So I assume i'm on the "first" BIOS. No idea how to use Qflash so I don't really wanna go that route. Should I just download the BIOS file from the Gigabyte website and run it?

As for the chipset, since I have the gigabyte ones installed now, can I just install the AMD chipset drivers over the existing ones?

 

[SacraficeMyGoat]

1) The F12 beta versions are for sure much more reliable. Unless you can't find a solution I'd skip F20a, just consider it last resort.

2) Disable Dual BIOS from the switch on the board

3) Best method is on a USB stick and use Q-Flash from BIOS itself, quick and reliable

4) Better you update to the latest from AMD

5) Except BIOS, same stuff





It works without Hybrid for me, also on F20a.
Never been able to make it 100% reliable with Hybrid even with F11.
It's just a few watts more power consumption but more reliable and quicker to resume. I'd go without.



Spread spectrum is adaptive.
It doesn't usually go below 99.6; you have some strong EMI noise around the board.

Okay, i'll try F12a first, thank you!

Do I need to format the USB before putting the BIOS on it, or can I use one with data on it? After the BIOS is on the flash drive, I just enter the BIOS and press the QFlash button on the back?

Also, my replacement 3700x came today so i'll be installing that tomorrow. Will I need to download the drivers again or?

 

[SacraficeMyGoat]

The USB flash drive needs to be FAT32, so you probably have to format it. UEFI can only read up to FAT32, it's unaware about more advanced filesystems.

Thank you! I just downloaded the AMD x570 chipset drivers; previously had the Gigabyte one.

Will download the BIOS tomorrow.

 

[SacraficeMyGoat]

Hey guys, so I just got done installing my replacement CPU (still 3700x.) I also installed a Corsair H150i Pro XT rgb along with iCue software. Also updated to F12g BIOS from F11.

1) Do I need to download chipset drivers or any other drivers since it's still a 3700x?

2) Every time I open HWinfo64 now, my PC freezes and I have to hold the power button down and reset it. Anyone know what's going on?

 

[SacraficeMyGoat]

Welcome to the f** club on number 2. I have the hard freeze issue as well while starting several programs (and A LOT while loading HWinfo). It's been doing it a lot less lately, but it still does it from time to time. I'm now struggling with constant idle crashes (12 overnight!) and made a new thread in the AMD General section.


So far I have not found any solution to work. Seeing as you have the same issue, I'm really tempted to order myself a new board from a different brand. It is defnitely nearing chucking the whole thing out of the window and buying some Intel hardware.

Try disabling Link/Asetek in HWINFO under safety settings. Seems to have helped.

 

[SacraficeMyGoat]

1. no

2. Is it only HWiNFO 64 that is showing this bizarre behavior? If so try uninstalling and reinstalling or better yet use the portable version that runs without need of being "installed"

I think it was only with HWiNFO. It seems to conflict with Corsair iCUE and I changed a setting to disable link something or other, and it seems fine now.

 

[SacraficeMyGoat]

Would someone more experienced than myself care to look at my HWiNFO Idle log and tell me if anything looks off?

I just installed my replacement 3700x yesterday along with a Corsair H150i RGB Pro XT AIO and the idle temps seem to fluctuate between 37.5C to 49.5C.

I'm not able to upload the CSV file format apparently, so if anyone could tell me how to upload the file, that'd be appreciated!

Edit: Files posted below!

 

[SacraficeMyGoat]

Just zip it and upload the archive.

Thank you.

 

[SacraficeMyGoat]

You have to fix something with the Power Saving plan.
The single Core X VID voltages are going down to 1.013v.
But the CPU Core VID Effective is fixed at 1.4V and Voltage SVI2 TFN almost always at 1.4V.
Mines are dropping both below 1.0V in idle.

Take 1usmus Universal Plan and modify it accordingly to this post:
https://www.reddit.com/r/Amd/comments/eng1y8/ryzen_3700x_high_iddle_temps_and_fan_speed/

Test again and see if they drop.
Don't seem like you have Global C-States enabled, C6 residency is at 0%. Any reason?
Doesn't seem you use the PBO EDC Bug.

If you don't use/want C-States set the minimum processor state in the power saving plan to 30%.

View attachment 355546

Thanks for taking a look. I'll try editing the power plan settings and upload another log in a bit.

As for Global C State/PBO EDC, I'm not quite sure what those are and they're likely disabled? I haven't done much configuring in the BIOS aside from XMP and a couple of small tweaks for stability. I'm still learning my way around AMD software (first time having an AMD machine.)

Should I enable Global C State? I assume this setting is in the BIOS?

What is the PBO EDC bug?

Sorry for all of the questions; Coming from Intel to AMD has been quite the change. Although I never did much tweaking with Intel aside from general OCing/Undervolting.

 

[SacraficeMyGoat]

If you didn't touch it, Global C-States should be in Auto and enabled.
It's in Tweaker>Advanced CPU.
Check first the power plan, maybe it's there the issue.
If you don't see the C6 residency going above 0% in idle try to force it enabled in the BIOS.

You can check the EDC bug here:
https://www.overclock.net/forum/13-amd-general/1741052-edc-1-pbo-turbo-boost.html

But if you are not experienced I'd wait a little bit before venturing in it.

Alright, I set the custom power plan, and checked the BIOS and Global C State was already enabled.

It looks like C6 Residency is still not going above 0% in idle. I attached another log below with the custom power plan enabled at idle.

Do I need to enable C6 in the BIOS?


Edit: Package C6 Residency seems to stay at 0% while under load as well. I assume this isn't good?