Hexus review up - GTX480

[SlowSpyder]

Why are people comparing a single gpu to a dual gpu?.....dual gpu's have issues single gpu's do not for starters....and as afaik the 5970 still gets raped in tesellation by the 480?

Same reason others compared the GTX295 to the 5870.

Dual GPU's do have potential scaling issues, but this is AMD's answer on the highest single card, Fermi is Nvdia's. Both are DX11. The prices line up as expected it appears (5870 -> GTX480 -> 5970) with performance. Apparantly the 5970 uses the same amount of power (if that review is correct). So, if someone wants a high end card they'll have to compare the potential trade offs, but there's no reason they shouldn't be talked about together.

 

[GaiaHunter]

I've seen graphs of cards in Dirt2 with DX10 and DX11 cards together combined: http://www.xbitlabs.com/articles/video/display/radeon-hd5830_11.html#sect1

How can you conclude 100% that DX11 was not used for DX11 cards and DX10 for DX10 cards?

http://www.xtremesystems.org/forums/showpost.php?p=4307989&postcount=1986


As you can see they have graphs 4 Dirt 2 and 4 Dirt 2 DX11

2560x1600 4xAA Ultra DX11

GTX 480 - 50.69
5790 - 63.7
5870 - 47.71
5850 - 40.34

 

[SolMiester]

Why not. issues used to be bad scaling and power consumption (285 vs 4870x2)
The scaling is still there but not as bad. Tesellation isn't a game and I could say the same about bout Crysis. I wonder how loud the 480 is.

If it weren't for the immense power consumption, I'd say its a good buy.

Saying tesellation isnt a game is a bit naff mate, as it could very well prevail is the coming games. I wouldnt wont to spend $600 and have to upgrade for games 6mths later...

 

[MegaWorks]

Why are people comparing a single gpu to a dual gpu?.....dual gpu's have issues single gpu's do not for starters....and as afaik the 5970 still gets raped in tesellation by the 480?

If AMD decides to lower prices on the 5970, then considering it over the GTX480 seems reasonable to me don't you think?

 

[*kjm]

You care about MPG in NASCAR?

If powerconsumption, not performance, is your thing...go mid/lowrange....and quit whining

Faster is faster....get over it

Corsair HX620
Asus P7P55D-E
Intel 860
8 gig Gskill
5870
ect...

Mid to low range?

 

[SolMiester]

If AMD decides to lower prices on the 5970, then considering it over the GTX480 seems reasonable to me don't you think?

I dont do dual GPU setups!

 

[cbn]

Saying tesellation isnt a game is a bit naff mate, as it could very well prevail is the coming games. I wouldnt wont to spend $600 and have to upgrade for games 6mths later...

In other threads people have said tessellating a low quality model out to high quality is going to save a lot of memory bandwidth compared to rendering a high quality model without tessellation.

So it is likely ATI will bump up tessellation with HD6xxxx.

 

[Lonyo]

Saying tesellation isnt a game is a bit naff mate, as it could very well prevail is the coming games. I wouldnt wont to spend $600 and have to upgrade for games 6mths later...

Check out the Dirt 2 figures.
When you enable DX11, the gap changes from being 44% at 1920x1200 with 4xAA in DX9 to 6% with 2560x1600 4xAA and DX11 (GTX480 beating HD5870)

If you think the GTX480 scoring 2x or more the HD5870 in extreme tessellation in Heaven means much, then that's fine, but from the limited results here, it doesn't always translate into anything meaningful game wise.
Having an advantage which shrinks dramatically when DX11 is enabled in a game which uses at least some tessellation and then saying "woo it's super fast in a benchmark" doesn't really work.

Hopefully with some more reviews and hopefully some in depth testing the impact of DX11 and various DX11 features will become clearer (Dirt2 I think allows specific DX11 features to be enabled), so it'll be a case of waiting before saying that a tessellation benchmark has any relevance to actual experiences.
Metro 2033, AvP and Stalker are others which will be useful.

 

[Skurge]

I've been meaning to ask, what other DX11 effects does heaven have except tessellation?
If the 480 looses ground to the 5870 in DX11 games (dirt2 and BC2) with different effects and only destroys it in a benchmark that makes extreme usage of one single effect, it kinda makes you wonder about its true DX11 performance.

 

[JSt0rm]

This is what Keys said he couldn't stop smiling about, eh? A

Well I would be smiling if I got 3 of them in the mail for free

 

[blanketyblank]

http://www.xtremesystems.org/forums/showpost.php?p=4307989&postcount=1986


As you can see they have graphs 4 Dirt 2 and 4 Dirt 2 DX11

2560x1600 4xAA Ultra DX11

GTX 480 - 50.69
5790 - 63.7
5870 - 47.71
5850 - 40.34

I guess that means in real world games the tessellation really does slow down the 480 a lot. In the DX10 benchmarks Fermi looked to be dominating the 5870 (NOTE 10.2 catalyst so does not include the Dirt 2 performance boost from 10.3 which was pretty big I think ), but once in DX11 it's less than 3 fps difference which is like 5%. Even that could be and probably is attributed to the VRAM advantage since they just give the 2560 x 1600 scores.

 

[Stoneburner]

Charlie may have helped Nvidia this time around by setting the bar so low. The results are about what was expected. Of course, different sites may have different results. It seems there's never any "accuracy" with the results but there is "precision" in terms of where the numbers fall as a relative matter.

 

[Dark4ng3l]

Well I would be smiling if I got 3 of them in the mail for free

Maybe he just realized that this thing pretty much guarantees he will never have to waste electricity on heating ever again.

 

[konakona]

Thing is though, those titles don't stress a 5870 <1920x1200. So we really needed a boost in the heaviest games, where it didnt deliver.

Otherwise, good performance

my sentiments exactly! will need more reviews to come to any conclusion though.
15&#37; oc isn't bad, not sure how well it scales (yes I did see those figures, but would like to see a more comprehensive comparison between oc'd fermi and oc'd 5850).

You could argue most people don't oc, but this is an enthusiast territory anyway. Oc'ing does matter if there are significant gains to be had, as in an oc'd 5850 being faster than stock 5870.

 

[psychoticdream]

http://hardforum.com/showpost.php?p=1035509550&postcount=166

another bench out

Official IXBT.com benchmarks.

Test Box:
Intel Core I7 CPU 920 (2667 MHz);
ASUS P6T Deluxe;
3 GB DDR3 SDRAM Corsair 1066MHz;
WD Caviar SE WD1600JD 160GB SATA.
Tagan TG900-BZ 900W.
Windows 7 32bit; DirectX 11;
Dell 3007WFP (30").
CATALYST 10.3; NVIDIA 197.17.

Game: Unigine Tropics Benchmark

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-tr-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-tr-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-tr-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-tr-wxp-2560-pcie.html

Game: Far Cry 2

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-fc-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-fc-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-fc-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-fc-wxp-2560-pcie.html

Game: Unigine Heaven Benchmark DirectX 11.0

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h2-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h2-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h2-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h2-wxp-2560-pcie.html

Game: CRYSIS, RESCUE

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr3-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr3-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr3-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr3-wxp-2560-pcie.html

Game: CRYSIS WARHEAD, CARGO

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr4-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr4-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr4-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-cr4-wxp-2560-pcie.html

Game: Unigine Heaven Benchmark DirectX 10.0

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h1-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h1-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h1-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-h1-wxp-2560-pcie.html

Game: 3DMark Vantage Graphics MARKS

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-m-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-m-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-m-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-m-wxp-2560-pcie.html


Game: Colin McRae: DiRT2

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-di-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-di-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-di-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-di-wxp-2560-pcie.html


Game: Just Cause 2

Result:
1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-jc-wxp-1280-pcie.html
1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-jc-wxp-1680-pcie.html
1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-jc-wxp-1920-pcie.html
2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-jc-wxp-2560-pcie.html

 

[ural]

You care about MPG in NASCAR?

If powerconsumption, not performance, is your thing...go mid/lowrange....and quit whining

Faster is faster....get over it

Are you getting paid to run your videocard? What kind ass-backwards comparison is that? NASCAR teams have sponsors that PAY for fastest cars and their drivers. Do you have such sponsors? I don't.

Besides, the general trend in technology is that more and more performance is achieved at higher and higher efficiencies. For example, newest car engines are more efficient (consume less gasoline) at delivering the same or even better output. HD 5870 delivers about 50&#37; more performance compared to 4870 at the cost of only 20% more power consumption. Not to mention HD 5870 actually consumes about 20% less at idle than 4870.

GTX 4XX series seem to be completely out of this general trend in technology.

 

[T2k]

http://hardforum.com/showpost.php?p=1035509550&postcount=166

another bench out

Just checked Dirt 2 (no DX11 for NV) numbers... seriously: no AA & no AF and suddenly GTX480 beats 5970 and GTX480SLI beats everything... can a reviewer any more patently biased (= "paid" in Russia)?

 

[blanketyblank]

What does 3d mark vantage actually test? Strangely enough it seems 5870 beats 480 at all of the resolutions in that review although we know 480 should be the faster card. At least I know ungine stresses tesselation, but what is vantage doing?

 

[psychoticdream]

guy posted one more

9. Warhammer 40.000: Dawn Of War 2
, No AA, No AF
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1280x1024 http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-1280-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1680x1050 http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-1680-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1920x1200 http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-1920-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 2560x1600 http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-2560-pcie.html
AA 4x + AF 16x
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1280x1024, AA+AF http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-aaa-1280-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1680x1050, AA+AF http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-aaa-1680-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 1920x1200, AA+AF http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-aaa-1920-pcie.html
&#1056;&#1072;&#1079;&#1088;&#1077;&#1096;&#1077;&#1085;&#1080;&#1077; 2560x1600, AA+AF http://www.ixbt.com/video/itogi-video/0310/itogi-video-do-wxp-aaa-2560-pcie.html

yeah i give up i'm not sure its properly done

 

[Grooveriding]

Congrats to all who waited 6 months to pay $100 more for what they could have had with 5-10&#37; less performance 6 months ago for $100 less.........



I'll let you know how it goes with my 6XXX series in November while you're waiting for the 580....

 

[Genx87]

Congrats to all who waited 6 months to pay $100 more for what they could have had with 5-10% less performance 6 months ago for $100 less.........



I'll let you know how it goes with my 6XXX series in November while you're waiting for the 580....

Sounds awesome, see you in November.

 

[schneiderguy]

http://www.guru3d.com/article/geforce-gtx-470-480-review/

 

[Hauk]

http://www.guru3d.com/article/geforce-gtx-470-480-review/

Damn, looks like we're all doin the same thing..

 

[schneiderguy]

http://anandtech.com/video/showdoc.aspx?i=3783

:awe:

 

[Schmide]

up

Edit: Stop hammering it and let an old man read.