The RX Vega 64 Owners Thread

[Davedree]

v56 trades blows with the 64 and the 1080 when put at the same clocks actually and also my v56 is running at around 180w at 64 speeds so.

Is that measured using power meter/clamp or just from gpu-z core only estimation?

 

[gupsterg]

As of now it is on par with 1080 and delivers similar performance. Though to be honest the power consumption of V64 would steer me more towards 1080 if it were not for Freesync. I am an AMD fan but I'm not so blinkered I would declare Vega 64 viable competition to GTX 1080 if you weren't tied to freesync.

+1. The other gripe currently I would think any potential buyer has is lack of AIB cards.

Just a quick scout on web reveals for ~£30-£40 more than OCUK PowerColor V64 you gain an AIB GTX 1080 without reference blower. If any ref cooler VEGA owner is reducing clocks/voltage to gain quieter experience on VEGA then TBH they've lost performance vs an AIB GTX 1080 IMO. This performance loss may or may not be consequential depending on what frame rates they're targeting. Plus if they have FreeSync I do believe the game experience will be better than GTX 1080 without G-Sync, which sorta nullifies targeting max FPS.

I do not think VEGA was as good as the hype. Besides the wait for launch (plus it's issues) people are still waiting for AIB cards. I would think many are probably dismayed by this.

Here's to hoping AMD really keep pushing out some great drivers and "we" get games which make use of the card well.

 

[nashathedog]

I'm on the LC card so can't respond directly but have you increased the fan RPM from the stock setting? I know on the LC card the temps can hit the max target without increasing the fan RPM...

No I haven't I'll turn it up and see how it acts today, Thanks.

 

[Davedree]

+1. The other gripe currently I would think any potential buyer has is lack of AIB cards.

Just a quick scout on web reveals for ~£30-£40 more than OCUK PowerColor V64 you gain an AIB GTX 1080 without reference blower. If any ref cooler VEGA owner is reducing clocks/voltage to gain quieter experience on VEGA then TBH they've lost performance vs an AIB GTX 1080 IMO. This performance loss may or may not be consequential depending on what frame rates they're targeting. Plus if they have FreeSync I do believe the game experience will be better than GTX 1080 without G-Sync, which sorta nullifies targeting max FPS.

I do not think VEGA was as good as the hype. Besides the wait for launch (plus it's issues) people are still waiting for AIB cards. I would think many are probably dismayed by this.

Here's to hoping AMD really keep pushing out some great drivers and "we" get games which make use of the card well.

+1 with your take on it all.
For 460-480 you can get an Aib 1080, and destiny 2 is Inc which if you want reduces the price to 420-440. Which then makes the 1070ti price points even more pointless.

 

[GoogalyMoogaly]

It's the same price as ~equal performing Nvidia cards, depending what you want to do with it it's a better/same/worse purchase, looks priced perfectly to me /shrug.

And that's fine, Vega is priced alongside a 1080 and so it should perform accordingly.

And the £700 Vega 64 LC edition is priced the same as some 3rd party 1080Tis. So I guess it matches that performance across most games too?

 

[ubersonic]

And the £700 Vega 64 LC edition is priced the same as some 3rd party 1080Tis. So I guess it matches that performance across most games too?

The AIO cooled Vega 64 is priced against the AIO cooled 1080, both are noticeably cheaper than AIO cooled 1080ti's.

 

[GoogalyMoogaly]

The AIO cooled Vega 64 is priced against the AIO cooled 1080, both are noticeably cheaper than AIO cooled 1080ti's.

But do either of them perform as well, for the same money?
Vega 64 LC is the only option for a non-blower Vega 64 at the minute, the 1080 and 1080Ti have lots of non-blower options.

 

[ubersonic]

But do either of them perform as well, for the same money?

Depends what you mean? If you're asking if the AIO Vega and AIO 1080 perform ~ the same for the same money then yes, if you're asking if either of them perform as well as an air cooled 1080ti then no ofc not.

 

[LtMatt]

And the £700 Vega 64 LC edition is priced the same as some 3rd party 1080Tis. So I guess it matches that performance across most games too?

The £640 RRP RX Vega 64 AIO, which i bought three of at that price from different retailers, cost the same as 1080 AIO's at the time.

 

[JediFragger]

cost the same as 1080 AIO's at the time.

With the 1080 AiO's boosting to 2Ghz out of the box...

 

[silversurfer]

Seems like a pointless argument but in that example I'd say the correct comparison is the 1080ti AIO because of design, noise factors is part of what people are buying. Its pointless because pricing and stock availability varies by region, seems they are short of stock but thats not about the card itself and this is the owners thread not the discuss buying one thread so its a bit misplaced as a criticism

 

[LtMatt]

With the 1080 AiO's boosting to 2Ghz out of the box...

That would be expected though wouldn't it? Air cooled Vega sits around 1450Mhz, LC sits around 1700Mhz.

 

[henson0115]

Is that measured using power meter/clamp or just from gpu-z core only estimation?

thats what the psu says, so measured.

 

[JediFragger]

That would be expected though wouldn't it? Air cooled Vega sits around 1450Mhz, LC sits around 1700Mhz.

Yes, but when the the cards were first benchmarked then the Vega LC was barely able to keep with a 1080 FE, whereas the 1080 AiO would probably give another 10-15% performance on top of that. I suppose my issue is with pricing/performance when it comes to the LC/AiO.

 

[ICDP]

If any ref cooler VEGA owner is reducing clocks/voltage to gain quieter experience on VEGA then TBH they've lost performance vs an AIB GTX 1080 IMO. This performance loss may or may not be consequential depending on what frame rates they're targeting. Plus if they have FreeSync I do believe the game experience will be better than GTX 1080 without G-Sync, which sorta nullifies targeting max FPS.

Vega AC can maintain better than stock performance while consuming considerably less Watts over stock. Still not 1080 levels of efficiency but better much improved over the out of box experience.

https://forums.overclockers.co.uk/t...king-undervolting-information-guide.18793012/

I do not think VEGA was as good as the hype. Besides the wait for launch (plus it's issues) people are still waiting for AIB cards. I would think many are probably dismayed by this..

Vega is a decent GPU, nowhere near the hype but neither is it a total failure. Though 2x months and no AIBs is abysmal.

Here's to hoping AMD really keep pushing out some great drivers and "we" get games which make use of the card well.

I wouldn't hold out much hope, I believe the "AMD fine wine" meme is mostly unwarranted. Both AMD and Nvidia improve drivers and performance over time. There is also the fact that the last few drivers released from AMD for Vega owners have been quite poor. Losing overclocks/undervolts on almost every reset, game profiles not working, display corruption in GUI etc.

Overall I am happy with Vega and do not regret getting one (especially at the £450 release price).

 

[ICDP]

Depe

Yes, but when the the cards were first benchmarked then the Vega LC was barely able to keep with a 1080 FE, whereas the 1080 AiO would probably give another 10-15% performance on top of that. I suppose my issue is with pricing/performance when it comes to the LC/AiO.

Depends on the review, but for the most part I saw LC 64 slightly ahead of 1080 and AC 64 slightly below. Stock for stock LC V64 is about ~15% faster than the AC model.

 

[Davedree]

thats what the psu says, so measured.

OK fair, but I'd imagine the psu is only measuring current draw on the pcie power cables, not factoring in the pcie slot at an additional 20-30 watts if using the stock cooler fan. Or more if watercooling + twin fan rad etc.
But 200-210 watts is perfectly fine, just wish peeps would plot the frequency curve vs fps to show if the investment of aio liquid/or custom water is worth it for the gain in performance compared to undervolted or tweaked vega at it:s sweets pot.

 

[LtMatt]

I'd be interested to see if other Vega users can get the following clock, hbm and voltage settings stable, while using whatever the lowest amount of voltage is required to achieve stability. I'd be interested to know what voltage is required to achieve these clocks from AC and LC samples if possible.

With this particular Vega 64 XTX I'm using at the moment, i have achieved the following which is 24/7 stable.

Stock core clocks, but core undervolted 0.075mv. State 6 core clock increased to 1702Mhz.
HBM overclocked to 1070Mhz.


That results in a constant core clock between 1702-1752Mhz.

 

[henson0115]

OK fair, but I'd imagine the psu is only measuring current draw on the pcie power cables, not factoring in the pcie slot at an additional 20-30 watts if using the stock cooler fan. Or more if watercooling + twin fan rad etc.
But 200-210 watts is perfectly fine, just wish peeps would plot the frequency curve vs fps to show if the investment of aio liquid/or custom water is worth it for the gain in performance compared to undervolted or tweaked vega at it:s sweets pot.

considering the board is powered by the psu to, i imagine the figure is correct. i am undervolted to 980 i think. The figure sounds accurate to when you work it all out, its been a few weeks since ive done so; so going from memory.

 

[henson0115]

I'd be interested to see if other Vega users can get the following clock, hbm and voltage settings stable, while using whatever the lowest amount of voltage is required to achieve stability. I'd be interested to know what voltage is required to achieve these clocks from AC and LC samples if possible.

With this particular Vega 64 XTX I'm using at the moment, i have achieved the following which is 24/7 stable.

Stock core clocks, but core undervolted 0.075mv. State 6 core clock increased to 1702Mhz.
HBM overclocked to 1070Mhz.


That results in a constant core clock between 1702-1752Mhz.

at work atm so cant test but i reckon i could reach those as i can push 1700 on 1080 undervolt. although i am 56 and not 64.