[RFC] What compromise I should make on my motherboard purchase?

[nac]

Going by the google search, it's pointing to drivers. Not hardware.

When I installed drivers, I installed MSI Live Update 6 utility which scanned and updated some of the drivers/utilities (don't remember what are those drivers/utilities). Scanned for system information and chipset driver installed is 17.x, AMD has released 18.x few days ago. I may give it a try.

When I launched Bluestack for the first time, there was this pop up suggested to enable AMD-v which I did. Don't remember remember if the issue happened after I enabled or even before I enabled. Some of the suggested fix is to disable virtualization.

Win 10 Pro Version 1803 (17134.376). It's up to date.

PSU is about 2yrs old, it's part of the upgrade. I don't know if the voltages on the rails are stable. But my doubt is not on any hardware. If I can check that using any software, I can do that. If I have to do a test using muliti/amp meter or something, than NO. I don't know how to do that and I don't have a multi meter.

 

[quicky008]

use it with vt disabled for a while,check if the problem still occurs or not.

 

[nac]

There is a feature I can use, but coz of this issue I gotta diable? That's a bummer.

I upgraded to latest graphic driver Adrenalin 18.xx. I followed what AMD suggested, even though when I installed (of course after uninstalling old one), it said Crimson ReLive which is the older version. I checked the downloaded setup file, which says 18.xx. I don't know why it says the installed one is old but the one I installed is new. AMD doesn't seem to have an option to uninstall chipset driver alone and there is no way to check the version I installed in windows device manager other than using MSI utility which still says the installed version is 17.4xxx. But graphic card version is correctly displayed in Radeon software.

As of now, installed new graphic driver and still enabled virtualization. If the issue again pop up, I will disable AMD-v.

There is this feature which I am not aware of virtual resolution. This message popped up before, but I could read the message fully it disappeared few times. When it showed last time, it again disappeared before I could read the whole message. Out of curiosity I googled and read the whole message. Now I am using my FHD monitor as 2K display. Tried 4k, but it's too small to read. Now I am toying with this idea of dropping the idea of getting 4k display and get a 2k one. With this feature I can use it as a 4k monitor.

 

[nac]

Next BSOD.

This time it's akikmpag.sys

Out of 3 fix, 2 are already done. Third one is replacing the file. But I can neither copy paste the file nor rename the file. I tried changing the ownership from System to my user id and given full control to my user id. Nothing works.

 

[whitestar_999]

Always use DDU Display Driver Uninstaller Download version 18.0.0.2 to uninstall gfx drivers,it is much more efficient than standard uninstall(do read the instruction before).

Using lcd/led monitor at any resolution other than its native resolution is not a good idea.I still think issue is with win 10/driver conflict.

 

[whitestar_999]

Check this thread,it might even be an issue with ram brand(though I still don't believe it completely):
Video TDR Failure (atikmpag.sys) on Windows 10 | Community

 

[nac]

Yes, I used DDU latest version and in safe mode. Instruction suggests not to use DDU all the time, only when there are issues. Since I have issues, I used it.

So using 2k monitor as 4k is not a good idea... Okay.

 

[nac]

Tried OC genie from windows environment using MSI command center utility. It sets core @ 3.75ghz and fan speed @ 100%. This 6-7% core OC gives about 6-7% performance gain. This few seconds bench doesn't shoot up temperature any more than NON OC bench. Couple of degrees isn't much. Don't know if this is like Ryzen master which don't remember OC settings after shutdown. Here is the snapshot.

*i.imgur.com/PWWLNvj.png

Don't know why GPU always stays @ 400mhz whether idle or playing 4k video. It stays the same. Even 4k doesn't call for higher clock speed, what will? Don't feel like setting it other than AUTO in BIOS. Will check that this weekend.

 

[whitestar_999]

Don't OC unless you feel the need for it(aka you are hitting cpu usage above 70-80% regularly). As for gpu frequency 4k videos not necessarily stress gfx that much,try with a game.

 

[quicky008]

have you been able to stop the random crashes and bsods?

 

[nac]

Set GPU clock @ 1100mhz and checked there isn't seem to be any difference when playing 4k video. @ 1100mhz GPU always stay there like it did @ 400mhz, may be @1100 GPU usage chart doesn't show many peaks and troughs like 400mhz did. Average temp, graphic usage, graphic memory usage all are about the same with both 1100mhz and 400mhz.

Tried to push by running handbrake for about 4 minutes. CPU stayed below 50% and GPU didn't even hit that much, may be about 10-20%. So decided to run Passmark benchmark and ran bench. After 3-4 unsuccessful attempt I googled and learned that it's coz of dual monitor and running other app at the same time. Non-OC CPU score is about 5% less than the average and OC score is about average.

Compared with my old rig scores, performance gain in %
Single Multi
vs E2160 150% 490%
vs E8400 80% 280%

Don't OC unless you feel the need for it(aka you are hitting cpu usage above 70-80% regularly). As for gpu frequency 4k videos not necessarily stress gfx that much,try with a game.

Probably I could never need that. My CPU doesn't seem to hit that much and stays there for long. I run HWinfo to check the usage for the day, but because of BSOD (and also manual restart without remembering that I am running HWinfo) I don't know what's the average usage over an extended period of time.

have you been able to stop the random crashes and bsods?

I tried the suggested fix I could do. But still I experienced BSOD. So formatted the disk and installed OS from scratch and there is no BSOD for more than 24hrs. Since it's random I am keeping my fingers crossed.

 

[whitestar_999]

I just remembered something.I read some amazon.com review about a gigabyte ryzen mobo which has this strange issue of sudden crash which needed cmos to clear(bios reset) to start again & apparently it was caused by hwinfo running in background(it does interact with hardware sensors on mobo after all).

 

[quicky008]

This sounds rather worrying - are these crashes caused by some issue in the motherboard itself or the cpu?



Sent from my XT1068 using Tapatalk

 

[nac]

I just remembered something.I read some amazon.com review about a gigabyte ryzen mobo which has this strange issue of sudden crash which needed cmos to clear(bios reset) to start again & apparently it was caused by hwinfo running in background(it does interact with hardware sensors on mobo after all).

That's strange. I will try that if/when I see next crash.

This sounds rather worrying - are these crashes caused by some issue in the motherboard itself or the cpu?

No, it's a driver issue. Not hardware.

 

[nac]

Formatted and clean installation of OS fixed it, been using it for about 1.5 months now without that crash.

If anyone interested in load/usage/temp, here is the snapshot of HWINFO.

*i.imgur.com/lYDIXqf.png

I am not gamer, but recently installed PUBG/emulator and played. @ Balanced & Extreme graphic settings I am getting upto ~ 60fps (low may be around 30 fps, not so sure) and avg. temp is around ~58 C
@ HD and medium settings temp is over 60C and @ FHD and low/medium around mid 60s. Since I don't like to see temp over 60C, I settled for Balanced and extreme.

 

[quicky008]

which version of win 10 are you running at the moment?