and here it is, at stock values. https://imgur.com/UgZq9Yo
-
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
Sorry for my multiple posts, but seems that running on stock/UV only/OC & UV seems it has no real benefit in benchmark in games; tested ffxiv, division, sottr. It's probably because of my cpu bottleneck, however in 3dmark firestrike, where there is no bottleneck, it's 22k (stock) vs 25k (oc&uv). For the moment, till my new cpu will arrive, I will keep the gpu at stock values to see if I encounter any blackscreen.
Thanks for you long reply, as for the memory clock, I had is also at 1050Mhz/1000Mv and it also crashed. I did experiment with different values and all seem to crash, however I haven't tested with pure stock clocks/voltages. Yes, my psu is a m12.
To be honest, if I will not get this sorted out (OC & UV), the card will be probably RMA'd to the store.
I had tried applying your OC&UV settings, however, it stays like that only in desktop/idle, if I run firestrike or sottr benchmark, the clocks are usually lower for the gpu, however the memory one stays the same.
Here is a screenshot of my benchmark, anyway.
The clocks you put in wattman for core frequency will never be reached, it will always be 50-80mhz lower in the real world, that's just how Vega behaves. And yeah, it's a very clear bottleneck as you can see: GPU Bound 33%. This should be >90% ideally.
Definitely keep running stock in the games you had problems, see if it's stable, because if it's fine at stock then you can start pushing the card later to find what the limits are.
Soldato
The card has two independent feeds of the connectors right?
Soldato
so you do not have two independent feeds to the gpu, that is a bad idea...yes, however I use the same cable to provide power 2x 6+2 pins. I didn't try with a second power cable because I can't find the box of my psu
Soldato
yes, however I use the same cable to provide power 2x 6+2 pins. I didn't try with a second power cable because I can't find the box of my psu
This means you dont.
Which is the next thing to do before blaming any other component. Get even a molex from another line and use an adapter to see if this placates the blackscreens.
I got the question wrong it seems. As an adapter, you mean something like this? https://s12emagst.akamaized.net/pro...2783c2586dd60a0e0b07be2be3b1_450x450_pl8k.jpg
Soldato
Yes (molex to pci-e) sometimes they come with the GPU but would be the best way if you cannot locate the modular cables that came with your psu.
Definitely find the cables. That's probably why you black screen. I wouldn't use molex adapters tho, too many problems and they're usually dodgy. PSU is the one component you should absolutely not mess around with. If you can't find the cables then get a new PSU, this will be an issue for any card that requires more than one pci-e cable.
I remember this issue clearly at launch, a lot of people did the same mistake.
I called my mom and told her to search for the cables in our house and she did find them, the thing is, next week I will be able to go and grab them. Till then, I ordered today from an online store a molex adapter, will use that for 1 week or so to see if I have the same problem.
Also, another update, I got in touch with someone from a local forum who bought the same gpu from the same store a few days ago and he also has blackscreen issues, using only 1 cable to provide power (like me). The funny side, the store had a lot of vega64 nitro+ gpu's left and two other guys (as far as I know) bought the same model one week ago and they told us they use 2 cable to power and they don't have problems. In summary, this might be the problem, but we will see.
Thanks for you replies,
Also, another update, I got in touch with someone from a local forum who bought the same gpu from the same store a few days ago and he also has blackscreen issues, using only 1 cable to provide power (like me). The funny side, the store had a lot of vega64 nitro+ gpu's left and two other guys (as far as I know) bought the same model one week ago and they told us they use 2 cable to power and they don't have problems. In summary, this might be the problem, but we will see.
Thanks for you replies,
Last edited by a moderator:
Soldato
Hmm if its the power delivery then it may just be the particular PSU is not sending a clean feed. I have seen other vega users eradicate the issue when they changed PSU which would lead me to believe that its a combination of the Vega being fussy and the PSU not being that great.
See https://www.reddit.com/r/Amd/comments/8nnuya/psa_vega_black_screen_crashes/
and
The best explanation was with a Seasonic issue:
https://www.bit-tech.net/news/seasonic-warns-of-focus-plus-psu-compatibility-issues/1/
See https://www.reddit.com/r/Amd/comments/8nnuya/psa_vega_black_screen_crashes/
and
The best explanation was with a Seasonic issue:
https://www.bit-tech.net/news/seasonic-warns-of-focus-plus-psu-compatibility-issues/1/
The thing is, I can use my rx580 meanwhile. My RMA limit at the shop is 22 october. I will get the missing pci-e cable on 19, so I don't have very much time to test from there on. However, currently this issue happens daily a few times, so I guess the remaining time (3 days) might be fine to see if I have the problem still with the original cable. To be honest I'm really sad to return this.
The worst case scenario will be for 3 days to work without any problem and later having issues...
The worst case scenario will be for 3 days to work without any problem and later having issues...
What symptoms did you experience?
Did the screen briefly go black and then recover? Or did the whole PC crash?
Was it only when gaming or did you get random crashes while just using a browser etc?
My Seasonic 600w or 600w wouldn't run my Vega 64 but was completely stable after I upgraded to a Seasonic 700w. It's in my sons PC now and it's still rock solid. However I was getting random crashes with my Seasonic 850w and my Radeon VII (PC died completely so I suspect that was a motherboard issue just before it completely failed).
Whereabouts are you located? There may be a friendly forum member nearby who wouldn't mind trying it in their system to see if the card itself is stable on a different PSU.
Did the screen briefly go black and then recover? Or did the whole PC crash?
Was it only when gaming or did you get random crashes while just using a browser etc?
My Seasonic 600w or 600w wouldn't run my Vega 64 but was completely stable after I upgraded to a Seasonic 700w. It's in my sons PC now and it's still rock solid. However I was getting random crashes with my Seasonic 850w and my Radeon VII (PC died completely so I suspect that was a motherboard issue just before it completely failed).
Whereabouts are you located? There may be a friendly forum member nearby who wouldn't mind trying it in their system to see if the card itself is stable on a different PSU.
Last edited:
The monitor loses signal (black screen) , it happens in idle/browser and also in load/games, so it doesn't matter in the end. If I'm listening to music, it continues to play even if I have black screen (strange). The solution is to press the restart button.
I looked at the modular molex cable (where I did plug the molex adapter to 8 pin) and indeed, the end that goes in the psu connector has 6 pins instead of 8 (like pcie cable), so I don't think it was so wise to try, but hey, it was only 3 bucks.
I looked at the modular molex cable (where I did plug the molex adapter to 8 pin) and indeed, the end that goes in the psu connector has 6 pins instead of 8 (like pcie cable), so I don't think it was so wise to try, but hey, it was only 3 bucks.
Soldato
So it sounds like its not heavy load from your feedback, have you tried other drivers or changing the pci slot to test further?