RX Vega Owners Thread

Just wanted to add that as far as I know, the coil whine isn't actually harmful to the card if it isn't actually bothering you.
 
Got the vega 56 friday so had some time over the weekend..


1. Used wattman to undervolt and set the fan to run higher. I left the clocks the same for now as the ASUS is already factory OC. (1583 mhz)
2. Fan at 100% is quieter than the twin frozr 7950 at 100%. I play with headphones so that helps ignore then fan noise anyhow.
3. Played multiple games for several hours and runs pretty solid. No crashes.

Fallout 4 is roughly double the FPS with all the options turned up. (I had most of the options turned up before except for AA)
DOTA2 - FPS are higher but can't really tell it already ran fast. BUT it runs a lot cooler, only about 50C during this game)

I got it from amazon warehouse in "very good" condition for a good discount. Everything was there except the power cable, which I didn't need.
Did not see any scratches, dust, fingerprints, anything on he card itself. All the plastic caps were in place. Looks mint.
 
My admittedly limited experience with these cards has taught me this: Vega is SUPER SENSITIVE to power supplies.

Make sure that you are using different PCIe plugs from the power supply instead of using two plugs on the same cable, if you are doing that. Avoid using Molex-to-PCIe power adapters. Make sure your power supply is no less than 750W (again, in my limited experience), and you might STILL need to replace the power supply with a newer (reputable) one if it is more than a year or two old.

My Vega 64 whined like a bitch on my older Antec TruePower 850W. It was truly obnoxiously loud any time a game started up and nothing I did could get that to stop, so I replaced the power supply with another one I happened to have handy. It is almost entirely gone now using this much newer Antec TruePower 750W power supply. If I get into an older title that really ramps up the FPS, I can still hear it faintly, but like I said, mostly gone now. If that were still bothersome to me, I could actually enable the Frame Rate Limiter in the Radeon drivers to eliminate it entirely.

I haven't noticed any issues yet..

I'm using a Corsair HX620, yeah old school modular. I am using two seprate 8-pin's each going to their own plug on the power supply. HWinfo says the vega is sucking 244 watts max. I set the +50% on the power with the undervolt at 1050mv.
 
My admittedly limited experience with these cards has taught me this: Vega is SUPER SENSITIVE to power supplies.

Make sure that you are using different PCIe plugs from the power supply instead of using two plugs on the same cable, if you are doing that. Avoid using Molex-to-PCIe power adapters. Make sure your power supply is no less than 750W (again, in my limited experience), and you might STILL need to replace the power supply with a newer (reputable) one if it is more than a year or two old.

My Vega 64 whined like a bitch on my older Antec TruePower 850W. It was truly obnoxiously loud any time a game started up and nothing I did could get that to stop, so I replaced the power supply with another one I happened to have handy. It is almost entirely gone now using this much newer Antec TruePower 750W power supply. If I get into an older title that really ramps up the FPS, I can still hear it faintly, but like I said, mostly gone now. If that were still bothersome to me, I could actually enable the Frame Rate Limiter in the Radeon drivers to eliminate it entirely.

This is a new build I made myself for Christmas. Power supply is a Seasonic Focus Plus 750W that I purchased brand new. I am running off of just the one PCIe plugs that splits at the GPU, so I might dig up the box and run a separate cable for each 8-pin connector just to rule that out. Like I said it's not bad.. but if I can easily get rid of it by swapping out or adding another power cable i'll give it a shot.

Love the card otherwise.
 
This is a new build I made myself for Christmas. Power supply is a Seasonic Focus Plus 750W that I purchased brand new. I am running off of just the one PCIe plugs that splits at the GPU, so I might dig up the box and run a separate cable for each 8-pin connector just to rule that out. Like I said it's not bad.. but if I can easily get rid of it by swapping out or adding another power cable i'll give it a shot.

Love the card otherwise.
There was an article posted right here on HardOCP that stated that the Seasonic Focus line of PSUs had an issue when combined with the GTX 970 or Vega 56/64 GPUs and it was recommended the you avoid using that combo. Not sure if the problem is resolved or not but maybe the Focus PSU is a contributing factor.
 
There was an article posted right here on HardOCP that stated that the Seasonic Focus line of PSUs had an issue when combined with the GTX 970 or Vega 56/64 GPUs and it was recommended the you avoid using that combo. Not sure if the problem is resolved or not but maybe the Focus PSU is a contributing factor.


Seasonic claimed they only applied to FP units that were made before Jan off 2018 (and he said Christmas!) But I had the issue with 3 units, all made after March of 2019.

They finally "allowed" me to pay them $30 to upgrade to a new FP+ Gold 1000W. I don't have the VEGA s anymore but so far my VIIs are doing okay with it.
 
Should I get the Sapphire Pulse or the MSI Air Boost Vega 56?

I think all things being equal I would go with the Sapphire Pulse, but I can get the air boost for $260 and it comes with 2 free games, or I can get the pulse for $300. My current case has good airflow (Corsair air 540). Is it worth paying 40 dollars more for the Sapphire card and getting 2 free games that I would probably actually play(world war z and division 2)? How much louder is the air boost and how much more performance can I tune out of the Sapphire?


My case will actually be about 8 feet away from me because I have it hooked up in my living room to play on my Samsung q6fn tv. I need a freesync card because screen tearing is driving me nuts with my 1070 on this TV.
 
Should I get the Sapphire Pulse or the MSI Air Boost Vega 56?

I think all things being equal I would go with the Sapphire Pulse, but I can get the air boost for $260 and it comes with 2 free games, or I can get the pulse for $300. My current case has good airflow (Corsair air 540). Is it worth paying 40 dollars more for the Sapphire card and getting 2 free games that I would probably actually play(world war z and division 2)? How much louder is the air boost and how much more performance can I tune out of the Sapphire?


My case will actually be about 8 feet away from me because I have it hooked up in my living room to play on my Samsung q6fn tv. I need a freesync card because screen tearing is driving me nuts with my 1070 on this TV.

I have an Air Boost Vega 56 and a Powercolor Red Dragon (similar to your choices). The Red Dragon is obviously quieter, but I think if someone offered me another Air Boost and $40, I'd "downgrade" to it. I don't think you'll see $40 extra worth of performance (about 15% if my math is right) out of it even if you tweak it. The Air Boost isn't a bad card though if you don't mind the noise from the blower.
 
I have an Air Boost Vega 56 and a Powercolor Red Dragon (similar to your choices). The Red Dragon is obviously quieter, but I think if someone offered me another Air Boost and $40, I'd "downgrade" to it. I don't think you'll see $40 extra worth of performance (about 15% if my math is right) out of it even if you tweak it. The Air Boost isn't a bad card though if you don't mind the noise from the blower.
Do you think I will even be able to notice the sound? I sit about 8 feet away from the computer on the couch and play with fairly loud surround sound when im playing. How much performance would I be losing in comparison to my geforce 1070? I'm buying this card because the screen tearing is annoying me and I don't like using vsync. Here's a picture of my setup:

vresSLM.jpg
 
Depends on how loud you have your sound cranked up. I don't think it would really be any louder than turning on that fan in the corner. My point is really how much is $40 extra worth to you? All things being equal, get the Sapphire Card.

I think it's a pretty even trade between the two cards (1070 and Vega 56). Unless you are trying to play twitch shooters competitively, you should have a good experience with FreeSync turned on if the screen tearing was getting to you.
 
Depends on how loud you have your sound cranked up. I don't think it would really be any louder than turning on that fan in the corner. My point is really how much is $40 extra worth to you? All things being equal, get the Sapphire Card.

I think it's a pretty even trade between the two cards (1070 and Vega 56). Unless you are trying to play twitch shooters competitively, you should have a good experience with FreeSync turned on if the screen tearing was getting to you.
Edit: Nevermind, I re-read your post and you said 15 % price difference and not performance

By the way, the air boost version is actually the oc version. Not sure if that makes a huge difference.
 
Last edited:
So I'm going to necro this thread because I'm recently having some odd system crashing issues. I currently own a Vega64 with a Morpheus II aftermarket cooler. I have not adjusted any of the voltage settings, and have been satisfied with gaming performance by adjusting the Power Limit to +50%. This has been fine for almost every game I have played (Witcher 3, GTAV, Subnautica, Dishonored 1/2, etc.) with reasonably high graphics settings in the game. However, I recently purchased Black Mesa and after reaching the Xen levels, my system began to crash (straight to black screen, GPU fans turn on to 100%, graphics card appears to power off otherwise, motherboard LEDs/fans still running). I thought this might have been a fluke, so I just stopped playing the game. However, I recently went back to try out 7 Days to Die (a game I already have ~500 hours in) and it also crashes my system within a few minutes. After extensive research on this (and learning how damn unstable Vega64s can be), I have done the following:
  • Made sure I am using two separate 8-pin connectors from the same +12V rail
  • DDU'ed and tested 18x and 19x drivers.
  • Stress test memory, CPU, and GPU using AIDA64 (GPU stress test recreates the crash, see below)
  • Ran common games at ultra settings to see if I could recreate the crash (it does NOT happen in typical AAA titles, no matter how high the settings are)
Other important thoughts:
  • Running the GPU stress test in AIDA64 will immediately (as in, instantaneously) crash my system in the same manner if the Power Limit is set any higher than about -30%. I have tried initiating the test at -50%, running it for a few minutes, and then setting the Power Limit to 0%. Instantaneous crash.
  • All temperatures (running GPU-Z), under load, stay ~55-60C except the Hot Spot, which ramps up to 95-109C. This does not result in a crash, as that is the same hot spot temp I get in every game I have tested and maxed out the settings on.
  • PSU is eVGA 850 GQ, which is rated to be well in excess of the power requirements for the Vega64. The PSU is single rail, with more than enough overhead to cover power spikes.
Things I am pursing next to try to fix:
  • Ordered new Arctic MX-4 thermal paste. It should arrive this afternoon. I plan to clean and reseat the heat sink. I don't think temperature is causing this, but it is the cheapest first step to take.
  • If the above doesn't work, my next thought is that the PSU is dying (or for whatever other reason, doesn't want to play well anymore with certain games). I don't really understand this behavior, or why it is happening only in certain games, especially games that are less graphically intense than others. Replacing the PSU is the next least costliest step and I should probably have a spare anyways.
  • If replacing the PSU doesn't work, I'll simply create different gaming profiles that throttle the power in certain games depending on what is prone to crashing. I'm not in the mood to replace the graphics card itself (especially since I'm running an Omen 32 and the 5700 also appears to have crashing issues).
Thoughts on this are appreciated. Specs below:

Ryzen 7 1800x
MSI X370
Radeon Vega 64
Corsair Vengeance LPX (8gb x2)
eVGA 850 GQ
 
What about the newer 20.4.2 driver? It sounds like the black screen issues people were complaining about with older drivers.

You're not the first person to report issues on X370 boards (I think they were both MSI boards also) with 1st gen Ryzens that I've seen here recently. Newer bios available for the board? Memory pass memtest?
 
I don't have my Vega cards anymore. But I had some similar sounding problems with Nvidia and it turned out to be my RAM overclock.

Even though I was using the rated XMP profile, when RAM was clocked over a certain value (I believe it was 4133, so fairly fast) it would result in random crashes in 3D games. Could play for 2 or 3 minutes, or could instant crash. All normal desktop programs and tests went fine.

I never quite understood why that happened, but what I did was first remove both my CPU and RAM overclock back to stock speeds (RAM at 2133) and the problem went away. I then went in an added some juice to the RAM and was able to get back the OC.

In any case, I was tired of worrying about stability, and also heat/noise with the CPU overclock, and I've set everything back to stock and I haven't seen any game crashes in a while. Maybe that will help.
 
What about the newer 20.4.2 driver? It sounds like the black screen issues people were complaining about with older drivers.

You're not the first person to report issues on X370 boards (I think they were both MSI boards also) with 1st gen Ryzens that I've seen here recently. Newer bios available for the board? Memory pass memtest?
yea cause i have never had any kind of issue like neslepax is having...It could also just be his card is on the way out. I really dont think the performance is all that so much better with over 30% added power limit anyway. If dropping it below +30% fixed the issues i would call it a day.....sure a slight buzz kill but its still works within 5% of mine (if even that much) even if i have full +50% and full fan speed going. He should maybe try boosting the gpu voltage with afterburner and see what effect it has.
 
Yeah, boosting GPU voltage up sounds like a good idea (at least that worked for me on the system RAM).
 
I flashed BIOS to the latest version (did not fix the problem). I also ran memtest with no errors found. I have installed Afterburner. What is a good increment step to increase the gpu voltage by?
 
I flashed BIOS to the latest version (did not fix the problem). I also ran memtest with no errors found. I have installed Afterburner. What is a good increment step to increase the gpu voltage by?
try 1.2mv or 1.25mv..come to think of it...it can be done in driver performance settings as well even with out After burner (lol i totally forgot)
 
Last edited:
Fixed. It was the thermal paste. Even at -20% or -50% Power Limit, the Hot Spot temp under AIDA64 GPU test would regularly hit 100-109C. I am now able to push it to +50% Power Limit again, with the hot spot hovering around ~90C. If I keep the power limit around +20%, the hot spot drops to ~75-80C. Interestingly, I am also seeing a corresponding increase in the GPU power draw. My guess is that the "higher" power limits were drawing more power than could be adequately dissipated. I just didn't think it would happen so instantaneously and trigger an immediate shut down. I can't remember the last time I had to reapply thermal paste to a GPU, and I've owned a lot of flagship models over the years. Thanks for the advice, though! At least all of my driver files are current. ;)
 

Attachments

  • GPUZ pic2.JPG
    GPUZ pic2.JPG
    86.6 KB · Views: 0
Fixed. It was the thermal paste. Even at -20% or -50% Power Limit, the Hot Spot temp under AIDA64 GPU test would regularly hit 100-109C. I am now able to push it to +50% Power Limit again, with the hot spot hovering around ~90C. If I keep the power limit around +20%, the hot spot drops to ~75-80C. Interestingly, I am also seeing a corresponding increase in the GPU power draw. My guess is that the "higher" power limits were drawing more power than could be adequately dissipated. I just didn't think it would happen so instantaneously and trigger an immediate shut down. I can't remember the last time I had to reapply thermal paste to a GPU, and I've owned a lot of flagship models over the years. Thanks for the advice, though! At least all of my driver files are current. ;)

I just want to say: THANK YOU FOR POSTING THE SOLUTION TO YOUR PROBLEM.

So many people have some sort of issue they post about and then just vanish. It is not helpful for other people that may come looking for advice regarding similar issues in the future if there is no follow up from the original poster, even if that followup is a downer like "I never could get it right, so I replaced the unit."
 
I just want to say: THANK YOU FOR POSTING THE SOLUTION TO YOUR PROBLEM.

So many people have some sort of issue they post about and then just vanish. It is not helpful for other people that may come looking for advice regarding similar issues in the future if there is no follow up from the original poster, even if that followup is a downer like "I never could get it right, so I replaced the unit."

I totally understand that, especially after the past week of digging around forums for possible solutions to this problem and finding so many of them that had people never stating what they did after receiving helpful advice.

It's still curious to me that the crashes were only happening in indy games, though. I could easily force the crash in 7 Days to Die, a game that has been in perpetual alpha for years, but I couldn't get Witcher 3 or GTAV to crash at ultra settings. I'm not a software developer, so I don't know the answer to this question: does (I'm assuming) code that is not well optimized for GPU efficiency really push the hardware that significantly?
 
I totally understand that, especially after the past week of digging around forums for possible solutions to this problem and finding so many of them that had people never stating what they did after receiving helpful advice.

It's still curious to me that the crashes were only happening in indy games, though. I could easily force the crash in 7 Days to Die, a game that has been in perpetual alpha for years, but I couldn't get Witcher 3 or GTAV to crash at ultra settings. I'm not a software developer, so I don't know the answer to this question: does (I'm assuming) code that is not well optimized for GPU efficiency really push the hardware that significantly?
In a perfect world every game will push the gpu to 99% usage, yes in slightly different ways, but why one is crashing and others dont? To me thats more of a hardware defect and not on the devs....with that being said SOMETIMES were dealing with driver/firmware incompatibility which does happen as in the Black screens issues only some cards are effected
 
I totally understand that, especially after the past week of digging around forums for possible solutions to this problem and finding so many of them that had people never stating what they did after receiving helpful advice.

It's still curious to me that the crashes were only happening in indy games, though. I could easily force the crash in 7 Days to Die, a game that has been in perpetual alpha for years, but I couldn't get Witcher 3 or GTAV to crash at ultra settings. I'm not a software developer, so I don't know the answer to this question: does (I'm assuming) code that is not well optimized for GPU efficiency really push the hardware that significantly?
The home screens in indie titles typically allow the GPU to run wild. I've had NV and AMD cards physically squealing when playing the simplest of games because the frame rate was so high.
 
In a perfect world every game will push the gpu to 99% usage, yes in slightly different ways, but why one is crashing and others dont? To me thats more of a hardware defect and not on the devs....with that being said SOMETIMES were dealing with driver/firmware incompatibility which does happen as in the Black screens issues only some cards are effected
Been playing the newly released Mafia II Definitive Edition, using Vega FE, Ryzen 2700 and Dell 3440x1440p monitor. Was getting tearing since it was over 60 fps, so I used chill set with min and max at 60hz. Amazingly Chill holds it right at 60hz, just like Vsync but with no lag! Very impressed, anyways GPU usage varies between 75% to high 80%, game runs buttery smooth be it at 60fps but with very consistent frame times. Sometimes less is more :D as a note: When a GPU is at 99%+, it can introduce lag, stutter, pauses.

Have not played a game in awhile with the 3440x1440p monitor, finding that aspect ratio very good especially in Mafia II seeing more of the city while driving. This has to be my favorite aspect ratio for gaming.
 
The home screens in indie titles typically allow the GPU to run wild. I've had NV and AMD cards physically squealing when playing the simplest of games because the frame rate was so high.
That's a good point. Maybe a frame rate limiter would help.
 
Back
Top