-
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
- Thread starter Kaapstad
- Start date
Soldato
interesting you say that - my strix vega64 has been under a waterblock for coming up on 2 years now and the hot spot has always hovered around the 100 mark......never paid it much attention as it does what i want it to do but the system is due a flush and clean down soon so reading this has me thinking i should repaste.
when you say you mucked up the paste application - how so? and how did you improve it second time round?
Associate
Not enough paste first time round. Need to be liberal around the bottom half of the chip, then tighten the top screw of backplate first, then the bottom ones.
I'm using an Alphacool block on a Red Devil '64, but I'd imagine the issue is the same.
Obviously, you need to make sure your paste is non-conductive...
Soldato
Ah ok. Liberal you say, from memory I treated it much the same way as a cpu so probably didn’t put enough on. Will dig out the ole tg kryonaut ready for the drain and strip down.
Hey guys.
I've started getting this error message a couple of times a day when i'm just browsing the internet and such.
https://i.imgur.com/VidKMUf.png
I have no idea how to troubleshoot things like this so I've no idea if it's a driver problem or if my Vega 64 has developed a fault. The message box appears after my screen has blacked out for a second or two.
Any advice would be most appreciated. Thanks.
I've started getting this error message a couple of times a day when i'm just browsing the internet and such.
https://i.imgur.com/VidKMUf.png
I have no idea how to troubleshoot things like this so I've no idea if it's a driver problem or if my Vega 64 has developed a fault. The message box appears after my screen has blacked out for a second or two.
Any advice would be most appreciated. Thanks.
Running overclocked and/or undervolted? What PSU and what kind of PCIe cable/cables are you running to the card? tried a factory reset through AMD driver install without keeping user settings? any windows update happened?
PSU is an 850W EVGA (i forget the model name), separate cables rather than a split one, tried multiple driver versions and the card isn't terribly stable on stock settings and becomes even less stable when undervolted even a bit.
Personally, I think my card is just a bit knackered as I bought used from a mining rig but the price was super low so I cant complain.
Mining shouldn't in most cases do that unless the person doing the mining is an idiot. Mining cards are usually undervolted and/or underclocked on the core which would increase the lifespan in theory. I would look at your OS installation. I've had my fair share of issues with windows and its silly updates.
I've had my card since 2018 and it's been repasting a bunch of times but I think at this point, it's actually screwed. It has been hitting 85c constantly and has recently started just crashing the whole system. I just took it a part and sure enough, the paste has dried up and the spread was not great.
It's been repasted now and now it's sitting at 75c under load (with a fairly passing fan RPM of 40%), however the core clock is refusing to go anywhere near 1600. It won't even go as high as 1300. I'm playing Everspace 2 right now and it's sitting snuggly at 1100 with 99% util. Not sure what to do here.
It's a Red Devil, paste is MX-4 and the pads are APT2560. P6 and P7 voltage is 1050, I've tried up to 1150 but that just results in higher temps with zero changes in core clock. Power draw is at +50%. Hotspot highest is 105c, VRM temp is 84c (~10 from core, which might suggest poor VRM pads?).
Any ideas?
EDIT: worth nothing, the core clock issue persists even with stock settings on the card.
EDIT 2: Yeah card is running 72-75c in games now, core clock is erratic between 1100 and 1300. GPU black screens randomly, requiring a reset.
EDIT 3: Turns out I have a pre-2018 Seasonic Focus Gold+ PSU which I believe has issues dealing with transient load on high end cards. I've contacted them, since it's supposed to have a 10 year warranty anyway.
EDIT 4: At the expense of performance (until my PSU is replaced) I have reduced core voltage to 950mv, reduced power to -25%. Not had a crash in hours now. However, core clock is dire (900-1000) but this is enough to game at 60fps on medium graphics. I hope the issues I have been having is literally due to the PSU not delivering enough juice.
It's been repasted now and now it's sitting at 75c under load (with a fairly passing fan RPM of 40%), however the core clock is refusing to go anywhere near 1600. It won't even go as high as 1300. I'm playing Everspace 2 right now and it's sitting snuggly at 1100 with 99% util. Not sure what to do here.
It's a Red Devil, paste is MX-4 and the pads are APT2560. P6 and P7 voltage is 1050, I've tried up to 1150 but that just results in higher temps with zero changes in core clock. Power draw is at +50%. Hotspot highest is 105c, VRM temp is 84c (~10 from core, which might suggest poor VRM pads?).
Any ideas?
EDIT: worth nothing, the core clock issue persists even with stock settings on the card.
EDIT 2: Yeah card is running 72-75c in games now, core clock is erratic between 1100 and 1300. GPU black screens randomly, requiring a reset.
EDIT 3: Turns out I have a pre-2018 Seasonic Focus Gold+ PSU which I believe has issues dealing with transient load on high end cards. I've contacted them, since it's supposed to have a 10 year warranty anyway.
EDIT 4: At the expense of performance (until my PSU is replaced) I have reduced core voltage to 950mv, reduced power to -25%. Not had a crash in hours now. However, core clock is dire (900-1000) but this is enough to game at 60fps on medium graphics. I hope the issues I have been having is literally due to the PSU not delivering enough juice.
Last edited:
WIthout being 100% certain it does sound like your PSU is the culprit. Black screens and reduced clocks usually is a sign of power issues. Remember to use 2 seperate PCIe 8 pin cables from the PSU instead of a splitter. Vega can easily fry a cable otherwise, happened on my platinum 860watt psu from corsair. 2 individual power cables solved the problem and I didn't have any black screens since then.
Yeah I read somewhere that this can actually solved the problem entirely, unfortunately I don't seem to have another 8 pin connector, so either it never came with one or I have misplaced the cable. I am hoping they just replace my PSU. It's been a few years so I am not holding out on it.
Well Seasonic got back to me. They are just going down the "That PSU is not powerful enough for a Vega64" line, which is fair since AMD recommend a 750, but that's assumed most people use a ****** PSU. This one isn't ******, it's gold rated from a reputable brand. So that sucks.
So update on my situation. Seasonic eventually agreed to replace my PSU. It arrived on Friday and I have connected it all up but I still have issues.
The whole black screen issue is gone, since the PSU no longer trips. However, my GPU is just not sucking enough power. Running Furmark @1440p is only drawing 250w (I believe it should be 320-350) and my clock speeds are barely going to 1200. Running independent 8 pin connectors as well. OC bios switch is selected on the card and the power limit is +50%. Temps are not even hitting 75c.
Not sure what to do now
.
EDIT: annnnd disregard everything. Still black screening whenever something spectacular happens on the screen. GPU is running at full clock speed, Radeon software was doing some adaptive clock speed thing to maintain 60FPS, so if I was getting 60FPS at 1000mhz it would just run at 1000mhz.
I am just going to buy a 1000w PSU and be done with it. Can't be bothered working within the bounds of the bare minimum.
The whole black screen issue is gone, since the PSU no longer trips. However, my GPU is just not sucking enough power. Running Furmark @1440p is only drawing 250w (I believe it should be 320-350) and my clock speeds are barely going to 1200. Running independent 8 pin connectors as well. OC bios switch is selected on the card and the power limit is +50%. Temps are not even hitting 75c.
Not sure what to do now
.EDIT: annnnd disregard everything. Still black screening whenever something spectacular happens on the screen. GPU is running at full clock speed, Radeon software was doing some adaptive clock speed thing to maintain 60FPS, so if I was getting 60FPS at 1000mhz it would just run at 1000mhz.
I am just going to buy a 1000w PSU and be done with it. Can't be bothered working within the bounds of the bare minimum.
Last edited:
Did they give you the same model seasonic? If so then it won't solve the problem as they had that fundamental flaw which made them no good for power users.So update on my situation. Seasonic eventually agreed to replace my PSU. It arrived on Friday and I have connected it all up but I still have issues.
The whole black screen issue is gone, since the PSU no longer trips. However, my GPU is just not sucking enough power. Running Furmark @1440p is only drawing 250w (I believe it should be 320-350) and my clock speeds are barely going to 1200. Running independent 8 pin connectors as well. OC bios switch is selected on the card and the power limit is +50%. Temps are not even hitting 75c.
Not sure what to do now
EDIT: annnnd disregard everything. Still black screening whenever something spectacular happens on the screen. GPU is running at full clock speed, Radeon software was doing some adaptive clock speed thing to maintain 60FPS, so if I was getting 60FPS at 1000mhz it would just run at 1000mhz.
I am just going to buy a 1000w PSU and be done with it. Can't be bothered working within the bounds of the bare minimum.
Soldato
Corsair or Superflower 1000W and be done with it 
Final update on this.
I just replaced the RMA Seasonic PSU (GX650) with a Corsair RM850 2021 and low and behold, problem still exists. Card is never going past 77c and it's still black screening and fans are ramping up to 100%. If I undervolt the card to 950mv and -30% power limit, it never happens, but my clock speed suffers greatly and maxes out at 1350mhz.
I'm bored, card is now going to be listed on some auction site (will make a great mining card, since it still works at their desired power limits) when my 3080 arrives. Sad state of affairs
I just replaced the RMA Seasonic PSU (GX650) with a Corsair RM850 2021 and low and behold, problem still exists. Card is never going past 77c and it's still black screening and fans are ramping up to 100%. If I undervolt the card to 950mv and -30% power limit, it never happens, but my clock speed suffers greatly and maxes out at 1350mhz.
I'm bored, card is now going to be listed on some auction site (will make a great mining card, since it still works at their desired power limits) when my 3080 arrives. Sad state of affairs
Did you try 2 separate PCIe cables coming from the PSU i.e. not sharing 1 cable with the 2 connectors?
Also have you tried the bios switch for power save mode and in AMD wattman/profile - using profile/balanced or profile/power saver