• 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.

Poll: ** The AMD VEGA Thread **

On or off the hype train?

  • (off) Train has derailed

    Votes: 207 39.2%
  • (on) Overcrowding, standing room only

    Votes: 100 18.9%
  • (never ever got on) Chinese escalator

    Votes: 221 41.9%

  • Total voters
    528
Status
Not open for further replies.
Running a 300+w card with a 550w GPU could definitely be a problem. But this is an issue that AMD have created for themselves. since no official reviewer with a system of appropriate specs is releasing reviews, everyone will instead be following guys like this. If the drivers are also not mature enough then we can expect to see lots of disapinting benchmarks.


vega was already delayed for months purely for driving maturity, hard to see how much more they can get out of this.

fwiw a half decent 1080ti breaks 31k or so in firestrike, a 2100mhz titan can crack nearly 34k.

a GTX 1080 cracks around this score, maybe a few hundred points less.

so basically exactly where the doom demo put it, around 10% faster than a 1080 in amd optimised titles, otherwise on par with a 1080.
 
OK fair enough I admit that I was at wrong due to the FE prices. The 1080 was still faster, and the 1070 was faster and cheaper.


Do you expect a lower end Vega to be faster and cheaper than a FuryX?

Yea personally i do. Fury X performance is sometimes 390x like at 1080p so i think it's more than doable.
 
Running a 300+w card with a 550w GPU could definitely be a problem. But this is an issue that AMD have created for themselves. since no official reviewer with a system of appropriate specs is releasing reviews, everyone will instead be following guys like this. If the drivers are also not mature enough then we can expect to see lots of disapinting benchmarks.

Pcper have got one on order. And drivers are bound to have some issues with the core being brand new.
 
The chap with the Frontier Edition is running a 550W PSU. Could be power throttling as he states the frequency is all over the place.
It could be thermal throttling, unable to maintain the peak 1600MHz Engine Clock. His case looks a bit cramped, cpu well cooled but not loads of airflow.
 
So we are back to the AMD are doomed status again then? :D


I'm happy with 1080 performance for the right price but we still have 1-2+ months left until the gaming cards are out so quite a bit could change in that time especially with drivers.....

Like I said before, my main concern is more the price now, with the mining craze and everything else putting all GPU prices up, I just can't see vega coming out as cheap as what people are wanting/expecting :(
 
Something is off with that benchmark. I can't find any normal FS scores for LN2 Fury X at 1450 but it matches a gtx1080 at max watercooled/air clocks in Graphics score using the extreme bench. With another 150 mhz on the core plus architectural advancement's Vega should be nearing Titan P performance. If not then something is very off with these cores and AMD have taken a step back with there architecture as even Fury X at 1600mhz would get some where in between OC 1080 and Titan oc.
 
Last edited:
I'm happy with 1080 performance

I hope its a good bit faster than a 1080 - I'm perfectly happy with my 1070 performance but already eyeing up upgrades with medium to longer term thoughts for some of the games coming in about 6 months time and I'll definitely be going for something a good bit faster than a 1080.
 
vega was already delayed for months purely for driving maturity, hard to see how much more they can get out of this.

fwiw a half decent 1080ti breaks 31k or so in firestrike, a 2100mhz titan can crack nearly 34k.

a GTX 1080 cracks around this score, maybe a few hundred points less.

so basically exactly where the doom demo put it, around 10% faster than a 1080 in amd optimised titles, otherwise on par with a 1080.
Which is really weird because they've had 6 months plus of driver development time since then.
 
Which is really weird because they've had 6 months plus of driver development time since then.

obviously they were hoping to get more power out of the gpu by optimising drivers, but clearly that's not the case, 6 months of driver work apparently has gained them 0% performance.

since the doom demo put it (in non amd optimised titles) dead on with a gtx1080.

and this release driver puts it dead on with a 1080, hell a highly clocked 1080 cracks 25k graphics score, this is equal with a 2.1ghz gtx 1070
 
Something is off with that benchmark. I can't find any normal FS scores for LN2 Fury X at 1450 but it matches a gtx1080 at extreme watercooled/air clocks in Graphics score. With another 150 mhz on the core plus architectural advancement's Vega should be nearing Titan P performance. If not then something is very off with these cores and AMD have taken a step back with there architecture as even Fury X at 1600mhz would get some where in between OC 1080 and Titan oc.


The thing is AMD have themselves stated typically clock speed of the air cooled FE is 1382MHz. If you use that clock speed then this Vega score is not wildly out. E.g., AMD made 25% efficiency gains but the drivers need to worked on and result in 25% less performance than expected.

Giant heap of salt needed and I dn't put any weight in this 3Dmakr score,however even AMD say not to expect 1600MHz in typical use case. AMD also have the Pro vega WX9100 clock speed at 1200MHZ IIRC, so soemthign in the 1300-1400 for gaming Vega in typical use but able to sometimes boost higher is not too far fetched.
 
Status
Not open for further replies.
Back
Top Bottom