Just had a close look at this
To get that graphics score the card would have to be 88% faster than one of my 290Xs @1000/1250 stock, this is not going to happen.
That score most likely is a doctored 290P CF score.
-
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.
390X or 980ti - which one are you going for?
- Thread starter Combat Fighter
- Start date
Just had a close look at this
To get that graphics score the card would have to be 88% faster than one of my 290Xs @1000/1250 stock, this is not going to happen.
That score most likely is a doctored 290P CF score.
Here is a 3dmark11 Extreme score by Matt using 290P CF if you knock off the big overclock on the cards you get about the same graphics score as in that bogus leak.
Score 9665, GPU 290P @1270/1625, GFX Score 9418, Physics Score 11988, Combined Score 9825, CPU 2700k @5.2, LtMatt - Link
Score 9665, GPU 290P @1270/1625, GFX Score 9418, Physics Score 11988, Combined Score 9825, CPU 2700k @5.2, LtMatt - Link
Just had a close look at this
To get that graphics score the card would have to be 88% faster than one of my 290Xs @1000/1250 stock, this is not going to happen.
That score most likely is a doctored 290P CF score.
Its just over 70% according to this.
4791 + 71% = 8192.
I think its doable if:
Even a Tonga Based 290X with HBM and Texture Compression should be 50% faster.
Its just over 70% according to this.
4791 + 71% = 8192.
I think its doable if:
Even a Tonga Based 290X with HBM and Texture Compression should be 50% faster.
Don't use that as it is overall score you need to compare just the graphics scores to compare GPUs.
I just want a 'real' next generation card, i.e. on 16/14nm.
28nm is very old now (although it also means it's a very mature process), so we're just seeing the limits of what's possible with old-ish technology rather than getting new technology
According to Techpowerup, the GTX 980 has:
45% more performance per watt than a GTX 780 at 1920x1080
47% more performance per watt than a GTX 780 at 3840x2160
102% more performance per watt than an R9 290X at 1920x1080
85% more performance per watt than an R9 290X at 3840x2160
(these are of course average gaming power consumption per performance, and not full-power compute performance)
So, in theory, it should be possible for AMD to double the gaming performance of the R9 290X. And realistically they may well be able to add 60% without too much trouble.
But it would be very very nice if we actually see a single 16/14nm GPU from either company that gives a tad more performance than an R9 295 X2 for ~£400.
28nm is very old now (although it also means it's a very mature process), so we're just seeing the limits of what's possible with old-ish technology rather than getting new technology
According to Techpowerup, the GTX 980 has:
45% more performance per watt than a GTX 780 at 1920x1080
47% more performance per watt than a GTX 780 at 3840x2160
102% more performance per watt than an R9 290X at 1920x1080
85% more performance per watt than an R9 290X at 3840x2160
(these are of course average gaming power consumption per performance, and not full-power compute performance)
So, in theory, it should be possible for AMD to double the gaming performance of the R9 290X. And realistically they may well be able to add 60% without too much trouble.
But it would be very very nice if we actually see a single 16/14nm GPU from either company that gives a tad more performance than an R9 295 X2 for ~£400.
Don't use that as it is overall score you need to compare just the graphics scores to compare GPUs.
Well Matt's overclocked 290 (None X) CF score that you linked has a 21% higher Graphics score.
Last edited:
Don't use that as it is overall score you need to compare just the graphics scores to compare GPUs.
isnt it fair to assume its all done on the same platform because its on the same chart? Otherwise who ever made this is a dork
Last edited:
Its about the same Graphics score as 2 947Mhz 290 None X running in CF mode.
With Tonga and all the stuff they could add to it i don't think that's out of reach, and its consistent, the 7970 was about as fast a 2x CF 6950's.
With Tonga and all the stuff they could add to it i don't think that's out of reach, and its consistent, the 7970 was about as fast a 2x CF 6950's.
Soldato
- Joined
- 27 Feb 2012
- Posts
- 6,586
So, what you're saying in short....
It's probably a fake score and they used a CF 290p to fake the score? Like what Kaap said, a few moments ago?
It's probably a fake score and they used a CF 290p to fake the score? Like what Kaap said, a few moments ago?
Well Matt's overclocked 290 (None X) CF score that you linked has a 21% higher Graphics score.
I also pointed out that you needed to remove the overclock first to compare the graphics scores. If you do this it looks like that err leak is just a pair of 290Ps running a mild oc.
The other thing in the screenshot that gives it away as fake is the fact that the card series is named. It is possible for futuremark to do this but normally they are so slow when new stuff comes along that it is probably not the case.
Associate
- Joined
- 13 Apr 2013
- Posts
- 1,107
- Location
- Scunthorpe
tbh if i was buying out of them two, i would by the one that performs best in the games i play and the up and coming releases. that is all that matters to me as i get piddly squat enjoyment out of benching.
I did remove the overclock. with it removed it comes in at around stock 290 (None X) CF.I also pointed out that you needed to remove the overclock first to compare the graphics scores. If you do this it looks like that err leak is just a pair of 290Ps running a mild oc.
The other thing in the screenshot that gives it away as fake is the fact that the card series is named. It is possible for futuremark to do this but normally they are so slow when new stuff comes along that it is probably not the case.
Last edited:
isnt it fair to assume its all done on the same platform because its on the same chart? Otherwise who ever made this is a dork
Review sites when posting bench results post overall scores which include physics and combined scores (big contributors to the score in 3dmark11). This makes it totally unfair and impossible to compare overall scores from a different PC like the one in the bogus leak.
On the other hand if you compare graphics scores in isolation it gives a far better comparison.
I did remove the overclock. with it removed it comes in at around stock 290 (None X) CF.
Which is pretty close to what that bogus leak got.
I think they are just running a pair of 290Ps for that screen shot with mild overclocks.
Which is pretty close to what that bogus leak got.
I think they are just running a pair of 290Ps for that screen shot with mild overclocks.
Just because its around the same score does not automatically mean IT IS 290 CF, if we had seen 7970 leaks priour to release you would have said that is just two 6950's.
290X just two base 7950's.....
Last edited:
Soldato
- Joined
- 27 Feb 2012
- Posts
- 6,586
Lol.
I'll stick with Kaap's thoughts, much more sensible.
Humbug will now continue to do weird maths that use 10%+15%-12% = something arbitrary = probably almost 100% correct score
I'll stick with Kaap's thoughts, much more sensible.
Humbug will now continue to do weird maths that use 10%+15%-12% = something arbitrary = probably almost 100% correct score
I know.
Always the one trying to drag the thread into the gutter.
Grown-up's are trying to have a discussion about hardware, please do the rest of us a favour, take yourself and your Teenage moodies to the park swing, brood about life over there, out of our way.
Or Keep it ontopic, instead of making it about people and attacking them. If you have a problem with my calculations point-out what's wrong with them.
Last edited:
Bogus leak Screenshot
2 of my 290Xs running stock
If you compare just the graphics score and also the individual graphics tests GT1, 2, 3, and 4
The difference in performance is almost identical to comparing 290P to 290X CF.
I think this leak is well and truly busted they are using 290P CF lol.
We all want the 390X to do well but come guys it is not going to be nearly twice as fast as a single 290X.
2 of my 290Xs running stock
If you compare just the graphics score and also the individual graphics tests GT1, 2, 3, and 4
The difference in performance is almost identical to comparing 290P to 290X CF.
I think this leak is well and truly busted they are using 290P CF lol.
We all want the 390X to do well but come guys it is not going to be nearly twice as fast as a single 290X.