• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

New Vega 56 Pulse Crashes

Associate
Joined
29 Jul 2016
Posts
42
Hello Guys,

I have random crashes during The witcher 2 game in 4k...
My card undervolted and running at P7 1652MHz 1025mv (1520-1560 MHz during gameplay), memory at 930Mhz (Hynix) 900mv (I`ve tried P7 1652 and 1000mv and 920Mhz HMB, same error)
GPU temps 71, HBM2 77 max, GPU VR VDDC - 81 max, GPU VR MVDD - 81 max, power draw 183w max
After 10-20 min I have Fatal Error massage: D3D Device was removed due to DXGI_ERROR_DEVICE_RESET, this usually indicates a driver bug, or GPU hardware failing/overheating. If you were running a game at the time of this error then game may have triggered a driver bug or caused your GPU to overheat...

My PC specs:
CPU: Ryzen 1600x at 3925MhZ 1.28v
Motherboard: ASRock Mini-ITX Motherboard - AB350 GAMING-ITX/AC
RAM: Ballistix Elite 16 GB Kit OC at 3200Mhz 1.35v
PSU: EVGA Supernova GS 550W '80 Plus Gold'
GPU: SAPPHIRE PULSE Radeon RX Vega 56 (bought it 1 week ago)

I`ve already reinstalled game and drivers... same error...

Here is screenshot of HWiNFO64 sensors: https://clip2net.com/s/3ZIbl6t

Any suggestions? What could be wrong?

Thank you in advance!
 
Associate
OP
Joined
29 Jul 2016
Posts
42
that's probably where the problem lies then.
try running your games with the side panel off and see if it still crashes.

I run The Witcher 2 with the side panel off + I underclock CPU voltage 1.28->1.1 at 3.8Mhz and GPU 1025 to 1000mv P7 at 1652Mhz
I was able to play longer but then I had the same error...
Here is screenshot of HWiNFO64 sensors: https://clip2net.com/s/3ZIqYeD
GPU temps before crash was 72, HBM2 78 max, GPU VR VDDC - 83 max, GPU VR MVDD - 83 max, power draw 186w max
After crash drivers will reset settings and GPU will draw 260w and fan on full speed

Is it PSU can`t handle the system power draw or overheating? but temps looks alright to me...
I don`t know what should I do next... new PSU 650 or 750w or new case?
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Does it crash at stock settings?

If you run the card on Balanced mode with 19.1.2 drivers, does it crash?
Did you use DDU to wipe everything before installing the drivers after you plug the new card?

No crashes on stock settings (balanced mode), but GPU will drop clock speed from 1550 to 1350...
I used DDU to uninstalled old drivers. I tried 19.1.1 and 19.1.2... same crashes when card i OC, but no crashes on balanced mode...

I was monitoring sensors and looks like this crashes related to overheating, because as soon GPU VR VDDC and GPU VR MVDD reaches 84C, then black screen and Wattman settings automatically drops to default... (1,5-2hr of gameplay)

Even on balanced mode GPU VR VDDC and GPU VR MVDD reaches 81-82, so it will overheat at some point on long run, because of the my small case and limited airflow...

I decided to change my case to Cooler Master MasterBox Q300L... I will keep PSU for now, to test card in new case with better airflow... Anyway I thought to change my case at some point, because I`m tired of how complicated is to change thermal paste or keep it clean inside...

I will update this thread next weekend with end result.

Thank you very much to all of you!!!
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Reduce P7 clock to 1632 and raise to 1050mv.

Raised P7 to 1050mv, GPU managed to keep clock speed at 1560-70Mhz and power draw was around 200-220w and it crashed after 45-60 min of gameplay... I`ve checked temps before the crash and GPU VR VDDC and GPU VR MVDD reached 87C, GPU temps was 73max, forgot to check HMB temps...
Also, usually I could go back to desktop, but this time PC rebooted itself...
Is it overheating or not? Maybe I need to change PSU to 650w?
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Go big, you already have a barely enough watts PSU.

https://www.overclockers.co.uk/ante...plus-gold-modular-power-supply-ca-24p-an.html

I feel your pain though, I'm running my V64 off of a 550w PSU as well atm while waiting for the new one. Pretty much throttle it myself to ensure it gets enough juice.

Thank you. I`ve already bought EVGA 750W gold!
I`ve tried to undervolt and underclock with HBM at stock 800Mhz and even with these settings it crashed...
Anyway, it was good excuse to buy new case and new cpu cooler, so I might be able to OC my 1600x to 4Ghz...
 
Associate
OP
Joined
29 Jul 2016
Posts
42
I run a 550w 5 year old Corsair PSU, my Sapphire Vega56 Pulse happily sits at 1630-1640, at about 60 degrees. Fan at 1300rpm.

Are we sure its a PSU issue and not just a problem with the timings/voltages?

Maybe it was issue with my OC settings. Unfortunately, I don`t have time right now to test it further...
 
Associate
OP
Joined
29 Jul 2016
Posts
42
I've changed PSU to EVGA 750w Gold and new case, It should be better airflow in CoolerMaster Q300L, but I still have that crashes every 1,5-2hrs of gameplay) My OC preset, P6 1582 1000mv, P7 1602 1025 and 930mhz on memory. Core temps 73 max, HMB temps 77 max...
I will try stock settings on balanced mode, to make sure its not the card itself cousing this crashes... if evrything okay, I will install new drivers 19.2.2 to see if I can have stable OC on new drivers... maybe its just the drivers... or I'm terrible in OCing )))
Error the same: DXGI_ERROR_DEVICE_RESET, this usually indicates a driver bug, or GPU hardware failing/overheating. If you were running a game at the time of this error then game may have triggered a driver bug or caused your GPU to overheat...
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Hello Guys,

The card doesn't crash on stock settings...balanced or turbo. It will crash if I do manual OC, auto-OC or auto-undervolt, but it is OC, so Overclockers guy said that if card is stable on stock setting, they can`t swap it.
I will have to keep it, so I was doing some tests and I found out that my OC card crashes only if core clock exceeds official boost clock of 1512Mhz...
I found out that card is stable on these settings: P6 1522MHz 1025mV; P7 1572MHz 1050mV and HMB2 at 930MHz, 900mV (I`m not sure if I can push mV higher to achieve higher MHz). Core clock seats around 1500MHz and I was able to play 5 hours without any crashes... not the best OC results but Im happy enough.

Thank you again everyone!
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Ahhh thought so. You need to lower your HBM overclock. The guys with HBM on the vega 56 forum have hit a brick wall trying to overclock Hynix. I would lower it and test in increaments until you find a safe overclock.

Thank you, I`ve already found stable OC: 1500MHz on core and 930MHz on HBM
 
Associate
OP
Joined
29 Jul 2016
Posts
42
Thank you, I`ve already found stable OC: 1500MHz on core and 930MHz on HBM

Ohh boy... My story continues... I thought I found stable OC but nope... I have random stutters when the card OC`d or on Balance mode... for example Dark Souls III: https://clip2net.com/s/40s7jFl or The Devision 2: https://clip2net.com/s/40s7l57 (both on Balance mode)

I`ve tried everything what I could... reinstall windows 10... changed power supply to 750 watt, better case with airflow... What I should do next? change motherboard and CPU just for GPU?))) any suggestions?

Also, I`ve sent it to Overclockers team but OC team send it back and said that everything is fine with the card (they tested in benchmarks in 1080p and 1440p).

I have no idea what to do next... I`m tired and I can`t enjoy my games, just waste of time... almost 2 months...

My PC specs:
CPU: Ryzen 1600x at 3800MhZ 1.15v
Motherboard: ASRock Mini-ITX Motherboard - AB350 GAMING-ITX/AC (BIOS version 3.60)
RAM: Ballistix Elite 16 GB Kit OC at 3200Mhz 1.35v
PSU: EVGA 750W '80 Plus Gold'
GPU: SAPPHIRE PULSE Radeon RX Vega 56
 
Last edited:
Back
Top Bottom