AnsweredAssumed Answered

1800x BSODs & Tolerances?

Question asked by chainik37 on Feb 24, 2018

Grabbed a 1800x recently & have had a run of BSODs on W10 x64, like plenty of others I guess.

Mobo auto OCs CPU up to 3850mhz, but running @3600, & 3700 don't show massive temp differences between these speeds +/- 2 - 5C;

This is combined with  -

Asus PRIME B350M-A (BIOS 3803) 

Team Group Dark Pro "8Pack Editon" 16GB (2x8GB) DDR4 PC4-25600C14 3200MHz Dual Channel Kit (Running @ 14-14-14-31 3200mhz);

ID-Cooling Auraflow 240mm


I have spent some time ensuring that the RAM is stable in Memtest but once in W10 this doesn't always seem to matter. 

Since getting the RAM 'right' my BSODs have all been core windows errors eg ntoskrnl.exe related faults.  Getting the RAM to this point has dramatically reduced the number & type of BSODs

(BTW in case anyone else is having a similar issue, fixing the RAM stability was only possible when I used AMD Ryzen Master to alter my CPU core Volts up to 1.35 & work from there.)

I cannot get past 20 min with Prime95 torture test without a restart/BSOD.  Temps are OK being reported as <70C when under 100% load by Ryzen Master.   Mobo temps never seem to get above 22-23C. 

Similarly AIDA64 won't run past 20mins without halting for an error.

I can run Handbrake for 30+ mins OK, tho this isn't hitting the system @100% load.


I am wondering about the values displayed by the monitoring s/w.

Both AIDA & HWMonitor show a value - Powers > CPU Package -Anyone know how this value relates to TDP / CPU tolerance?

I ask as e.g. I can see a temp value of 65-70C & a Package value of 116-133w when running a test.  This seems high to me when the TDP is 95w so I am wondering if I'm about to fry the CPU.  I don't get the relationship of these values yet...

What is the upper voltage tolerance?  I am peaking @ 1.406v max using the offset function to adjust it in real-time.  I assume a high voltage will kill the CPU quicker than a low one, & although this is a peak voltage draw I don't want to kill it for the sake of a small OC.


Ultimately I'd love the BSODs to stop & to get over the 1 hour mark for 100% stress tests but don't want to push too much until I know what these values should be to stay 'safe'.

Any info appreciated.