• 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.

The RX Vega 64 Owners Thread

Associate
Joined
5 Feb 2019
Posts
30
As I posted in another thread, I've been having a problem undervolting my reference Vega 64 (air).

If I first reset all the settings in Wattman, then start with a clean undervolt I can get my card stable with all states from p1 to p7 at 950mv. I can have Heaven running for a long time and HWinfo or GPU-Z shows GPU VDDC as 950mv, and the GPU Chip power is around 170W. The clocks will remain steady around ~1500mhz give or take a little. I leave HBM as is at 945mhz.

After exiting Heaven (or whatever game or benchmark I ran first) then try another benchmark or game then suddenly the GPU VDDC gets stuck at either 1050mv or 1100mv and the power consumption goes up to about 230 or 240W on average.

If I try changing voltages or clock speeds in Wattman or OverdriveNTool nothing effects that VDDC and it just stays at either 1050mv or 1100mv. Even resetting Wattman to defaults at that point doesn't do anything and the VDDC value stays the same. Reboots or shut downs don't do anything either and it's only until I force the card to crash by setting some really high clock that everything goes back to default.

Does anyone know what could be causing this, and if there is a way to prevent the core voltage (VDDC) getting stuck at a certain value?
 
Soldato
Joined
18 Feb 2015
Posts
6,480
P6 1537-1100 & P7 1632-1150 & HBM 1050-1100 +50% PL & Stock Fan Curve

Seems to hold 1600mhz in games unflinchingly with no more than 1400 RPM from fans - basically dead silent, certainly can't hear them over my intake fans. Could go all the way to 1100mhz HBM and it seemed stable enough in most benchmarks & games except for Odyssey, but since that's all I play lately, it had to go back to 1050. Also had success with 1075mhz but meh. Had some lower voltages & higher clocks OCs too but forgot them now, not too sure it would've been as stable anyway. Holds 40+ fps in Odyssey @ 4K Max (-clouds), which is not bad at all. Unfortunately the game has texture streaming issues which I can't get to the bottom of no matter what, but I guess Ubisoft's moved on now and put the team on the next AC. Was hoping for some HD Textures but alas, was not meant to be.

Forgot if I gushed about this or not but HBCC is love, HBCC is life. I just wish most games weren't so neutered texture-wise!
 
Associate
Joined
17 Dec 2018
Posts
35
Right so last night while playing Subnautica I got to a new area and my gpu went nuts.Usage went to 100% and the core clock climbed to 1670+ then crashed.Complete system crash in which I had to restart my PC.Tried that same area again and had another crash but this time just the game and driver cashed.So I ended up backing off 20mhz from my p6 and p7 core clocks and all seems fine now.

During benchmarking my pre crash settings with Unigine Valley and Superposition the card was stable and clocks stayed at around 1650mhz-1660mhz. Why did the clocks climb so much in Subnautica then?
 
Associate
Joined
4 Oct 2017
Posts
590
Location
Australia - Sunshine Coast
Right so last night while playing Subnautica I got to a new area and my gpu went nuts.Usage went to 100% and the core clock climbed to 1670+ then crashed.Complete system crash in which I had to restart my PC.Tried that same area again and had another crash but this time just the game and driver cashed.So I ended up backing off 20mhz from my p6 and p7 core clocks and all seems fine now.

During benchmarking my pre crash settings with Unigine Valley and Superposition the card was stable and clocks stayed at around 1650mhz-1660mhz. Why did the clocks climb so much in Subnautica then?
Could be anything. Problem is to track down the issue would be pretty hard remotely. Though likely an issue with the game and/or drivers.
 
Associate
Joined
17 Dec 2018
Posts
35
Could be anything. Problem is to track down the issue would be pretty hard remotely. Though likely an issue with the game and/or drivers.
You Know whats strange? outside of this one specific area those settings were fine and I had no issues,the core clock would peak to about 1650mhz as gpu usage went up and drop back down with usage linearly.I suppose this area realy pushed the card hard.
 
Last edited:
Soldato
Joined
18 Oct 2002
Posts
21,355
Location
Cambridge, UK
The odd thing is that's it's been working fine for 1 year plus :(

That's what lead me to believe it might be a driver/sofware issue but I really can't do much more then a fresh build of Win10 and the latest drivers which I've done.
 
Associate
Joined
18 Dec 2018
Posts
10
No, my monitor isn't a freesync monitor.
Double check your power cables the card. I thought my card was bad at first and then realized the 6+2 cable I had on one port wasn't completely connected all the way. It was crashing at the simplist benchmarks before I realized it the 2 side of the cable wasn't fully pushed in.
 
Soldato
Joined
18 Feb 2015
Posts
6,480
New PSU died but old one seems to be doing fine as well (OCZ 550w), though on a more conservative spec just out of paranoia (P7 1407-925mv, 0% PL, 945-900mv HBM). Not too shabby! HWinfo did report 242w Chip Power tho, which seems kinda excessive.
 
Soldato
Joined
18 Oct 2002
Posts
4,706
Location
Kent
Anyone have any recommended Wattman settings for a reference blower card (my water cooling gear wont be here for a few weeks)?
Thanks
 
Associate
Joined
1 Feb 2019
Posts
9
Hello all, I'm hoping someone can help me. I have just got my sapphire Vega 64 from here and am having some problems. I have it connected to a monitor via display port and a 4k tv via HDMI. I had this setup with my r9 390. Unfortunately when I'm using the 4k tv, as the primary monitor or duplicated, it keeps losing the signal. The will flash, or go black or show snow. It will then say no signal. It then flashes back to the desktop and off again.
I have tried different ports, settings etc. Turned off deep colour and tried different resolutions. It is a LG tv. I'm hoping there is a fix and it's not the card. Has anyone come across this?
Thanks.
 
Associate
Joined
7 Feb 2019
Posts
4
Just recieved my sapphire vega 64, and get the random black screen. 750 watt psu, 2700x no OC, 16 gb DDR4. GPU was hooked up with 2 cables from the PSU. It also hangs sometimes when the windows logo is on and the round loading bar just stops and freezes. Any ideas? I read somewhere that locking HBM frequency to 950 could solve it, but how to lock it when i cannot get into windows? Safe mode works though.

One time i got to install drivers load up a game and play it for 10 minutes. Then i restarted computer and it hang at the windows loading screen.

EDIT: Just want to mention, i had a 1060 that works perfectly. I removed the nvidia GPU drivers in safe mode with DDU.

EDIT EDIT: I might have forgotten to uninstall afterburner and rivatuner... Will try that later.
 
Last edited:
Associate
Joined
22 Feb 2009
Posts
217
Location
Tunbridge Wells
Hoping some folks here maybe able to help, i notice quite a few of us having driver issues.

I updated mine to 19.1.2 the weekend just gone and have had no end of issues with hangs and bsod. I previously had 18.11.2 installed with no issue.

I started to notice in BFV my gcard was getting to full load and fan speed where before it didnt even sweat. This was followed by the odd freeze and crash of wattman/adrenaline. The last few days i have tried about 5 or 6 drivers to try and eliminate these crashes.

Now I can no longer load onto desktop without it freezing or bsod as soon as Adrenaline loads on startup.

The only option is to go into safe mode, i have tried to fix registry, flash bios to latest, cleaned and reinstalled all evidence of amd drivers but to no avail. Even if i try old drivers that had previously been fine it will freeze at startup on desktop.

Any other options other than a clean install of windows?
 
Soldato
Joined
18 Oct 2002
Posts
21,355
Location
Cambridge, UK
Well it just shows you, I replaced my EVGA SuperNOVA 1600W with a Corsair CP-9020070-UK Professional Platinum Series 1200 and now more crashes!!!!

The final thing that gave me the the nod that it was the PSU was running OCCT, as soon as I run the "Power Supply" test it would cause a crash/reboot.

Thanks for everybody's help/patience
 
Caporegime
Joined
12 Jul 2007
Posts
40,434
Location
United Kingdom
Well it just shows you, I replaced my EVGA SuperNOVA 1600W with a Corsair CP-9020070-UK Professional Platinum Series 1200 and now more crashes!!!!

The final thing that gave me the the nod that it was the PSU was running OCCT, as soon as I run the "Power Supply" test it would cause a crash/reboot.

Thanks for everybody's help/patience
Good news and glad you debugged the issue. Always a pain these things.
 
Back
Top Bottom