New PC suddenly shutting down - R5 2600, Vega 56

[blingu]

Hi, I recently got my new computer. The specs are as following:

Ryzen 5 2600, OC to 4.0GHz (tested it several times)

RX Vega 56 Pulse

G Skill Ripjaws V 16GB 3200MHz

Patriot Burst SSD 240GB

1 TB Seagate mediocre HDD

Cryorig H7 Plus CPU Cooling

Corsair Vengeance 650M (PSU)

 

I've built the computer and installed basic stuff like drivers and programs that I usually use. I also downloaded Forza Horizon 4 and CS GO and played both of these games yesterday for a long time and haven't had any single problem with anything. Today I was playing Forza Horizon 4 and the computer suddenly just shut down that I even had to switch the PSU ON -> OFF -> ON so I can start my PC again. The best thing is, I didn't take part in races that could be challenging for my computer, I was just cruising casually around. Second time I was playing Forza again, did some races, cruised around and then - it shut down all of a sudden again. This problem doesn't appear in CS GO though.

 

The overclock that I did to my Ryzen 5 was tested by me yesterday and it worked perfectly fine. I also think it can't be an issue with the OC, because yesterday the computer was just running as it should. Everytime I start my computer there's a message that Radeon WattMan settings have been restored due to an unexpected system failure. I'm running my GPU on two powering cables.

[PlayStation 2]

What motherboard do you own?

[mxk]

Just because you don't think its the overclock doesn't mean you should disable it. Its usually a good idea to cross out of the possible problems.

[blingu]

I'm using the B450M MSI Mortar. I'll disable the OC.

[blingu]

On 12/16/2018 at 7:09 PM, mxk. said:

Just because you don't think its the overclock doesn't mean you should disable it. Its usually a good idea to cross out of the possible problems.

I reverted the whole OC, and it happened again (I found a hidden car on the map and it crashed, but it didn't crash in a challenging race with several AI cars). What I've noticed is before that happened I didn't have a Radeon WattMan message, but now after the crash it appeared there again. CPU OC seems to be fine.

[mxk]

20 minutes ago, blingu said:

I reverted the whole OC, and it happened again (I found a hidden car on the map and it crashed, but it didn't crash in a challenging race with several AI cars). What I've noticed is before that happened I didn't have a Radeon WattMan message, but now after the crash it appeared there again. CPU OC seems to be fine.

hmmmm thats really weird

[AngryBeaver]

If you have overclocked then you should know how to check our temps. See what your temps are while playing.

[blingu]

22 hours ago, AngryBeaver said:

If you have overclocked then you should know how to check our temps. See what your temps are while playing.

 

23 hours ago, mxk. said:

hmmmm thats really weird

I've played the game with the Radeon overlay to check the temperatures and usage of GPU. What I've noticed is in freeroam it's around 50-75% usage, in races 70-95. I've driven to the same place where my PC shutdown yesterday and got the car again (it didn't save due to the system failure yesterday). I've kept an eye on the usage, and it got dramatically high (max was 98% afaik). The computer shutdown instantly, no bluescreen nor anything, I had to switch ON -> OFF -> ON button of the PSU. The temperatures aren't the problem, they've been around 50*C to 69*C (didn't see 70 not even once). It's weird and I can't really understand the issue, especially finding a hidden car is just a cutscene and that's it. The radeon wattman settings restored message showed again once it crashed.

[blingu]

Just tried Furmark out, the PC shutdown after one second on 1280x960.

[blingu]

Can it have something to do with the PSU that has only one PCIe rail? Is it too little for the Vega 56?

[blingu]

I've read another answer in a video that the Vega card doesn't like 6+2 PCIe connectors but native 8. Is that true, may that be the problem? My PSU is a single rail, though it was in a high tier on the PSU list as far as I remember.

[Colb]

Have you ran your memory at stock setting (no xmp )  and just rules memory out .  My 2600x plus vega56 system for some reason is extremely fussy with memory and results in crashes exactly like your experiencing . Although I do now and again get a blue screen . However not always 

[blingu]

There's no reason why it should be the memory. Anyways, I rerun furmark again on 800x600, something in the case made weird noises, it weren't the fans (they went on at higher GPU temp). I don't know if it was the GPU or the PSU making that weird noises, but after the crash I tried to run furmark on 800x600 after some time again and it didn't even show up the benchmark video, it shutdown instantly.

[blingu]

What I've noticed is one of the Type 4 cables for PCIe has this red marked plug more inside than the others. I know that the left bottom plug should be like that, but both cables don't look exactly the same.

 

[reptillian_one]

On 12/19/2018 at 11:23 AM, blingu said:

What I've noticed is one of the Type 4 cables for PCIe has this red marked plug more inside than the others. I know that the left bottom plug should be like that, but both cables don't look exactly the same.

 

 

  Reveal hidden contents

 




 

 

hi Blingu,

 

any update on this? I've the exact same scenario, with the same CPU / GPU. Can sometimes game for up to 10 minutes then instant shutdown.

 

PSU is 650w Platinum, so figured it wouldn't be that..

[0x1e]

On 12/19/2018 at 11:23 AM, blingu said:

What I've noticed is one of the Type 4 cables for PCIe has this red marked plug more inside than the others. I know that the left bottom plug should be like that, but both cables don't look exactly the same.

 

 

  Reveal hidden contents

 

 

 

 

 

If you have to flip the PSU switch to get power again then there's a very strong chance you've tripped some kind of safety limit within the PSU.