Another Nvidia lie and trust you to take their word for it.
That graph does show how weak the AMD cards are at tessellation.
-
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.
CPR W3 on Hairworks(Nvidia Game Works)-'the code of this feature cannot be optimized for AMD'
- Thread starter tommybhoy
- Start date
That graph does show how weak the AMD cards are at tessellation.
It shows on the newer 285 though that they are addressing the issue, so hopefully the 390X should be fine with tess.
Agreed. It would be good for the AMD users to be able to get decent performance without having to lower quality with the CCC slider. Tessellation has been about for quite some time now as well, so quite surprised they are still that far behind.
It's probably like most of their 'features' - They will work on it for a short while, maybe get it in a game, add it to their spec sheet as a selling point then pay no attention to it ever again. It's the AMD way.
Not really
GTX 980 = 56376
GCN 1.2: 1796 Shaders @ 918Mhz = 31888
Theoretical.
GCN 1.2: 2816 Shaders (+57%) @ 1000Mhz (+8%) = 54069
GCN 1.2: 4096 Shaders (+45%) @ 1050Mhz (+5%) = 82320
It doesn't work like that AMD have fixed tessellation units, their new highend might well have the same number of tessellation units as R285 (which was double 290X).
Your being deliberate dumb, why do people do that? have they no self respect?
You said "That graph does show how weak the AMD cards are at tessellation"
The 285 is a GTX 760 level card and yet has similar tessellation performance as the GTX 770. clearly AMD's tessellation performance is not weak.
Compared with Nvidia it is if you pick out 2 to 3+ year old GPU's that have already gone out of retail production and are about to be replaced in literally weeks.
Your wrong, it works exactly like that, ask Kaap, he has all these GPU's, including the one that isn't listed there, the Titan-X, i predicted its score using the same formula and got to within less than 1% of Kaap's result. and yes i did that before i knew it.
Last edited:
Pretty sure Greg is doing it on purpose. 290x is old tech (2013), AMD just not released cards for a long time which is the problem. That will change soon with Fiji. Then he will likely start banging on about drivers
Soldato
- Joined
- 5 Sep 2011
- Posts
- 12,881
- Location
- Surrey
Do you even polymorph, bro?!
Who cares about Tessmark, it's a synthetic bench that doesn't really show any real world benefit (potentially) from NVIDIA's superior geometry engine.
Who cares about Tessmark, it's a synthetic bench that doesn't really show any real world benefit (potentially) from NVIDIA's superior geometry engine.
Pretty sure Greg is doing it on purpose. 290x is old tech (2013), AMD just not released cards for a long time which is the problem. That will change soon with Fiji. Then he will likely start banging on about drivers
Using sweeping statements like that? yeah i'd say so.
Pretty sure Greg is doing it on purpose. 290x is old tech (2013), AMD just not released cards for a long time which is the problem. That will change soon with Fiji. Then he will likely start banging on about drivers
Forget the Maxwell's then, what about the 780 laughing at the 290x ?
Anandtech must be wrong then? nope, don't think so Humbug you're just talking ****.
http://www.anandtech.com/show/8460/amd-radeon-r9-285-review/3
You can predict it on NVidia using that formula because of their Polymorph engine or whatever it is called. AMD have fixed units which is why they're so far behind on tessellation.
Soldato
- Joined
- 5 Sep 2011
- Posts
- 12,881
- Location
- Surrey
Here's one for the tinfoil crew, what if the GameWorks source code has data that could potentially tell AMD how to improve their geometry performance?
Whey more freebies. Funny how nobody thought of that? No?
Guess that will get glossed over as well...
Whey more freebies. Funny how nobody thought of that? No?
Guess that will get glossed over as well...
No they are not wrong, their results confirm the slide i posted.
Did you read it before you posted it? or are you just posting random crap claiming it agrees with you and hope no one reads it?
Last edited:
Soldato
- Joined
- 5 Sep 2011
- Posts
- 12,881
- Location
- Surrey
Talk about missing the point lol. Honestly, this is what you get for taking one statement I said and trying to turn it around with an outdated benchmark.
Good point. No idea. It is likely AMD will be addressing it soon though.
Did not even look at 780 (poor people who got that card struggling with new games at the moment it seems, pretty sure tessellation not on the top of their minds).
How many tessellation units does the 290X have then? then use your calculation to tell me how many the 290 has.
The answer is 4 for both of them because they are fixed units. 7970/280X have 2 and so do their chopped down versions because they are fixed units. Disabling and adding shaders has no bearing on how many tessellation units AMD cards have because they are independent fixed units.
Last edited: