Annoying AF crash error...plz halp

[SomeFknGuy]

I've been fighting an crash error for the better part of a year now with an older build that has been driving me crazy... I'll start with specs...

Ryzen 1800x, stock clocks stock volts
Crosshair6 hero mb
32gb corsair dominator(was previously 16gb G.Skill FlareX)
Radeon 6900xt(was previously Gigabyte Vega64)
Corsair AIO 360mm cooler
Corsair 1000w power supply(was previously 850 corsiar)
misc hard drives, games moved to SSD after crashing started and the spinny drive thrown in the trash because eff it.

I've been piece by piece replacing everything in this god forsaken computer trying to fix a crash that kicks the system to, practically, powered off. The MB kicks out error code 8 which according to reddit and asus is about the least useful in troubleshooting because it's the catch all "something is borked" message. Ram, PSU, GPU and PS have all been upgraded as time has progressed. I've nuked GPU drivers and reinstalled. MB and processor are all that is the same from when the crash issues began. At idle, web browsing, or watching movies it's solid as a rock. Playing games is where things go to shit. Temps on the CPU tend do peak around 65*c and the GPU around 70*c with an ambient temp in the room of 65*f. I've noticed that games based on the Unreal engine(pubG, Manhunter, and others) seem to cause the crash to happen much faster. MW2, the battlefield titles, warframe, overwatch, black ops, etc seem to all run fine until randomly the system shits itself...

Has anyone caught a crash like this before? My only real option now(i think) is to save up for a new mb/proc and hope for the best...

 

[TheHig]

Almost 100% of my Ryzen system crashes during some games have come down to ram settings. Not kidding- especially with first gen Ryzen.

Is your ram 4x8 or 2x16 ? Are you setting the docp / xmp setting to enabled?

What worked for me was setting the xmp/ docp (Asus boards ) and backing it down one speed step. For example if the ram is rated at 3200 - set xmp and manually back just the ram speed to 2933. None of my 1000 series Ryzen CPUs would do 3200 stable on 300 series boards.

Worth a shot of you haven’t done this yet. Alternatively set ram to auto and run it a default which is slow but if it’s stable gaming you know your on the right track.

 

[LukeTbk]

I would also look at the ram not able to run their XMP of the 1xxx-2xxx ryzen being a very possible issue and you just have been "unlucky" that it worked enough to let you boot versus the usual won't start at all.

You could obviously try after validating that the system is stable with the ram non overclocked default speed (is it 2666mhz ?), update your bios, play with voltage and change the used slot to see if they reach XMP, but running at the multiplier just lower after having set xmp could be the only way.

 

[SomeFknGuy]

Almost 100% of my Ryzen system crashes during some games have come down to ram settings. Not kidding- especially with first gen Ryzen.

Is your ram 4x8 or 2x16 ? Are you setting the docp / xmp setting to enabled?

What worked for me was setting the xmp/ docp (Asus boards ) and backing it down one speed step. For example if the ram is rated at 3200 - set xmp and manually back just the ram speed to 2933. None of my 1000 series Ryzen CPUs would do 3200 stable on 300 series boards.

Worth a shot of you haven’t done this yet. Alternatively set ram to auto and run it a default which is slow but if it’s stable gaming you know your on the right track.

Ram is 2x16, was previously 2x8. Didn't think about backing ram timings down... I've got xmp enabled and the ram is running at 3200mhz... Will try dropping it down. Trying another random thing...undervolting the GPU. If that fails to change it's behavior i'll back that ram timing down. This build only has to last me until the Zen4 version of the 5800x3d to come out and I'll be doing pretty much a rebuild of it all...

 

[GiGaBiTe]

Your CPU could have the dreaded segfault bug.

I was in your position back in 2017/2018 when I built a Ryzen 5 1500x system. Spent months trying to fix it when I learned about the segfault bug. Lo and behold mine had it, I RMA'd it back to AMD and got a working 1500x.

The behavior was similar to yours, light loads the system was fine *most of the time*, but any moderate to heavy load caused the machine to reboot/hard lock/shut off.

It's why I'll never buy any Ryzen 1000 series processor, you're walking through a minefield. They're long out of warranty and I doubt AMD would give you the time of day now for an exchange.

I thought the problem had been fixed, but earlier this year, I encountered a 3700x with the segfault bug on a customer's rig.