Is GTX 480 / Fermi a game changer for you?

[Steevo]

No, of cards period. DX11 doesn't matter right now, as that user base is still relatively small. DX9 and DX10 are still the most relevant APIs. Devs will still accept nv help because they have the larger user base. That's the point I'm trying to make.

Did you read the thread title by chance? Just wondering.


But with the logic you are showing, DX11 doesn't matter as Nvidia still pwns with total number of cards in the world so I will buy a Fermi as the 8800, er 9800, er whatever from Nvidia is great!!!!!


So lets never move ahead, lets never make anything new and better.


Is Fermi a game changer? No, it doesn't implement anything new, different, or with a large enough difference between the competition to make jack all of difference. 32AA is marketing gimmic, their ideas about their hardware and implementation have failed. At the uber high resolution people that have these play at Fermi is nothing more than a overpriced hairdryer.

 

[Dunceiam]

I missed this post , i'm gonna answer the Nvidia fanboy while it's best to ignore them i chose to not.

Obviously you missed the bottom of my post, reference it below.

(And to subdue the inevitable reply of, "You're just an Nvidia fanboy." I'll say it now, I've only owned ATI, my first card was a 4850, then a 4870, than a 4870x2.)

You're obviously incredibly biased with your opinion, and if anybody in this entire thread is a fanboy, it's you.

Anyway,

It is a dirty trick when you come up with some absrud 32XAA , it's not usable in games that put up some decent graphics , i wanna see you run crysis or recent metro with 32XAA , or even Batman that you mentioned , ohh you can't ? but why ? so in practice that feature won't be touched by most users , when i said 8XAA is overkill it was meant it doesn't make such a massive differnce , sure you can see it but you are playing the game or do you stare at jagged lines ? in games that pack some action you rarerly stop to look at lines , it's stupid and it's an argument a fanboy would give , look , Nvidia can do 100XAA , better than Ati , me stupid fanboy gonna take whatever bullshit Nvidia gives me.

The entire basis of your argument in regards to Anti-Aliasing is, "It requires a powerful card, so therefore it's useless." First of all, take a look at any GTX 470/480 benchmarks, you'll notice that when large amounts of AA are applied, the GTX 470/480 absolutely excel while older gen and ATI cards lack. To the point where the GTX 470 is performing better than the 5870. Nvidia's optimized Fermi to better handle AA, providing improved quality and performance, making large amounts of AA increasingly more accessible, by decreasing its effects on framerate.

You also make the argument, "It makes such a small difference, why do it?" (Granted, I've fixed a few spelling and grammatical errors). If the above was true, then there would be no such thing as audiophilia, nor would people buy high quality monitors (to use an obvious example). The contrast ratio increase between a low-end and high-end TN panel monitor doesn't give you a massive quality increase, only a subtle one. So according to your argument, why bother with a high end monitor if a low-end delivers a 'good enough' image?

And yes actually, I personally can see the difference between 8xAA and 16xAA, quite easily in fact, and I'm sure many other people can as well. Does this affect gameplay? No, not really. Does it reduce the entire immersive experience the intrinsically detailed graphics and textures served to create? Yep. Does it reduce the realism? Even more so.

Dear fanboy , GTS250 is a weak GPU who can't possible use 2 gb of memory , it can be helpfull for a HD5970 or a GTX480 but not for a GTS250 , if you don't understand this then the disscution ends here.

Dear Misinformed, the amount of onboard memory on a graphics card is not directly proportional to the power of the GPU. As mentioned before, the same amount of calculation goes on within the GPU, and therefore the same amount of memory will be used on a low-end card as on a high end card. The only difference between a high-end/low-end card in computational terms is simply how many times a second it can perform this task. Even if GPU power did correspond with the amount of memory, you'd also be wrong. The onboard memory on a graphics card is to store data, like a map in first person shooters, this allows the graphics card to avoid having to continually access much slower memory on the motherboard's RAM/HDD, and instead store the entire 'world' on the graphics card. Most loading screens in games are doing just that, they're buffering the GPU RAM and system RAM with game data, so the HDD doesn't continually have to be accessed. Memory has no correlation to how powerful a GPU is.

Also , for video edititing you don't require the memory to be on the video card , system memory is fine too , it's not like you have to run the video at 60 fps minimum and you need the data to be on the video cards memory so now swap , video content is easy compared to what games need from a video card and the speed to process it.

I can tell you've never coded anything in your life from the simple statement, "system memory is fine too." I don't want to turn this into a Computing 101 lesson, so I'll try to keep it short. Harddrives are incredibly slow compared to RAM, which in turn is slow compared to GDDR5, and so on. The generalized order of things in terms of speed is as such, HDD - RAM - GPU memory - L2 Cache - Shared Memory. It's a massive slowdown in a program running on a graphics card if it has to access global memory (RAM), which is what's lead to the invention of L1/L2 cache/Shared Memory. If you're running a CUDA program, that doesn't store any data on the graphics card's onboard memory, and instead only utilizes your system RAM, it's going to be incredibly slow. By increasing the amount of capacity on the card's onboard memory, you decrease the amount of times the card has to access system memory, and vastly increase the program's performance (in this case, video editing).

Oh, and if video content is so 'easy to render' compared to gaming as you claim. Then how come the average video renders at ~10 frames per second when we can game at 100+ fps? Explain that.

It's mindblowing what you say there , obviuosly you didn't touched video editing to see how cuda means nothing , actually Ati has a FREE plugin that speeds up adobe premiere and encodes with the gpu but it needs the whole system to be AMD , so praise Nvidia for not giving the same.

...you realize all Adobe programs offer free hardware acceleration, right? For both ATI/Nvidia.

Why do i care for those figures ? do every Nvidia fanboy split the atom ? research the cure for cancer ? build rockets ? sure sure , you have a rocket in your computer but you barelly run a game , i dont care how many billions operations Nvidia can do

Those are purely examples showing how much acceleration is capable by using CUDA. GPGPU computing is ideal for scientific purposes due to large parallelization capabilities, hence why most examples of CUDA performance are science orientated. However, that doesn't mean CUDA is limited to scientific applications. I'm also very curious to see how Fermi performs with Folding@Home.

why can't GTX480 beat HD5970 in games ?

>_>

Why can't a $10,000 car beat a $100,000 car in a drag race?

Besides, the HD5970 has two GPU's...the GTX480 has one. Hardly a fair comparison, no?

Again, I apologize for the long post and thread hijack.

 

[eidairaman1]

Im sorry the person who posted right before me, i couldnt get through a quarter of what he was saying before he bored me to sleep.

 

[leonard_222003]

Obviously you missed the bottom of my post, reference it below.



You're obviously incredibly biased with your opinion, and if anybody in this entire thread is a fanboy, it's you.

Do you think you are the first fanboy to lie you had every Ati card in history ? if we had to argue with every fanatic Nvidia or Ati fanboy that comes around we would grow white hairs on the balls.

The entire basis of your argument in regards to Anti-Aliasing is, "It requires a powerful card, so therefore it's useless." First of all, take a look at any GTX 470/480 benchmarks, you'll notice that when large amounts of AA are applied, the GTX 470/480 absolutely excel while older gen and ATI cards lack

Bla bla bla , so i guess you can play at full HD crysis with 16XAA ? or 32xAA ? not even 8xAA ? but where is the mighty power of GTX480 ? i don't think we need to go to metro 2033 , a game optimized for GTX480 to fully use it's resources.
[/QUOTE]

You also make the argument, "It makes such a small difference, why do it?" (Granted, I've fixed a few spelling and grammatical errors). If the above was true, then there would be no such thing as audiophilia, nor would people buy high quality monitors (to use an obvious example). The contrast ratio increase between a low-end and high-end TN panel monitor doesn't give you a massive quality increase, only a subtle one. So according to your argument, why bother with a high end monitor if a low-end delivers a 'good enough' image?

It depends on the person , some need monitor speakers and what a shock , most don't , some need a high end video card like 5970/5870/GTX480 and what a shock , most of them never consider buying a video card that go even over 100$.
AA is a good OLD way to improve image quality but at some point they got ridicolous with it , i give credit for Ambient occlusion , HDR , even Physx , i'm tired of absurd high AA modes , it's no innovation here.

And yes actually, I personally can see the difference between 8xAA and 16xAA, quite easily in fact, and I'm sure many other people can as well. Does this affect gameplay? No, not really.

So i guess you play all games at 8XAA and 16XAA ? at what resolution ? low end LCD 1280x1024 ? or high resolutions ? can you run at decent framerates metro 2033 with 8xaa ? or crysis ? how about 16xAA ?

Dear Misinformed, the amount of onboard memory on a graphics card is not directly proportional to the power of the GPU. As mentioned before, the same amount of calculation goes on within the GPU, and therefo

bla bla bla , long bullshit here
Dear disguised fanboy as Ati hardcore user , giving 2gb of memory to a weak GPU would be pointless because it coudln't process all the data in a requiered time , example a GTS250 would be framebuffer limited for a very high resolution so the brand integrator wold put more memory , if it can't process the data the memory wouldn't be much help , capishi ? :shadedshu

I can tell you've never coded anything in your life from the simple statement, "system memory is fine too." I don't want to turn this into a Computing 101 lesson, so I'll try to keep it short. Harddrives are incredibly slow compared to RAM, which in turn is slow compared to GDDR5, and so on. The generalized order of things in terms of speed is as such, HDD - RAM - GPU memory - L2 Cache - Shared Memory. It's a massive slowdown in a program running on a graphics card if it has to access global memory (RAM), which is what's lead to the invention of L1/L2 cache/Shared Memory. If you're running a CUDA program, that doesn't store any data on the graphics card's onboard memory, and instead only utilizes your system RAM, it's going to be incredibly slow. By increasing the amount of capacity on the card's onboard memory, you decrease the amount of times the card has to access system memory, and vastly increase the program's performance (in this case, video editing).

Oh, and if video content is so 'easy to render' compared to gaming as you claim. Then how come the average video renders at ~10 frames per second when we can game at 100+ fps? Explain that.

:shadedshu OMG !!! I'm gonna start to answer your last question , in video editing you don't render like in games , you render some effects (transitions , effects , menus ...) that don't need really FAST data transfers , that's why system memory is fine.Also , when you edit and render some effects you don't need realtime framerates and the need for a video card with huge quantity's of ram are not that crucial , system memory is good , video and photo editing , i'm not talking about other applications.
Sure more memory on the video card might help but not by much , when i render some effects i actually wait for the GPU to process a tiny 30mb file , now tell me wich is the one to blaim here ?

...you realize all Adobe programs offer free hardware acceleration, right? For both ATI/Nvidia.

Like i said , you didn't touched video editing software , it offers some extremely limited acceleration , sometimes it doesn't make a difference if you enable or disable , in after effects it's opengl and i didn't seen much acceleration but i hear adobe creative suite 5 will offer proper acceleration , premiere pro is so slugish i don't even use it , encoding is the pain and amd offers a plugin that speeds up the process , i was wrong , Nvidia too has a plugin for encoding but it costs a lot of money , it's called RapiHD.

Why can't a $10,000 car beat a $100,000 car in a drag race?

Besides, the HD5970 has two GPU's...the GTX480 has one. Hardly a fair comparison, no?

Again, I apologize for the long post and thread hijack.

Hehe , so HD5970 is a drag race car and GTX480 is just a 10k car ? this is how you see it ?
Saying HD5970 has two gpu's might be an advantage against GTX480 who has one , well the prices are the same , GTX480 consumes just as much and it's louder too , it's hoter too , it behaves like a dual GPU so no forgiving the beast.
I guess comparing GTX480 with HD5870 is fair , HD5870 is cheaper , it's much cooler and quiet , has 1 billion less tranzistors , consumes way less power and it's just 10-15% slower , is this a fair comparison for you ? yeeee Nvidia is the king.
Now before you get agressive with my posts , english is not my native language , i know i know some americans and englishmans think we all should know english very very well or die but tough luck , i don't and i'm lazy to use online spell check , when you people who point fingers learn chinese , french , romanian ( my language ) then you can point the finger at stupid stupid people like me who butcher the english language.

 

[Wile E]

I'm talking about some games that accept physx and Nvidia's whole deal to punish Ati , if you look at succesfull games that really make money they have nothing to do with physx , they make the game for everyone to enjoy with relative average/good hardware , examples :
WOW , warhammer 2 , BC2 , dragon age origins , TF2 , GTA4 , COD MW2 , i don't see physx in these games or Ati being excluded or running the game badly compared to Nvidia , these are smart developers who don't give a fu.ck about Nvidia's agenda.
On the other hand we have some games that pass trough our computers very briefly like crysis , crysis warhead , cyostasis , metro 2033 , these games have absurd high requirements for their respective times of launch and don't really make money like COD MW2 or BC2 , i could call them tech demo's but they are not that.
I'm talking about developers who made these games that will think twice before accepting Nvidia's help , how much money do they get from Nvidia to lose a lot of potential customers (ati) who has the hardware to run the game if properly optimized ?
Only one game was succesfull that had physx too , Batman , i bet you they lost some Ati costumers because they included physx.
I see the present mainstream something in the range of HD4870/HD4890/5770/GTX260/GTX275/GTX285 , who do you think has more market share here ? in the high end i see 5850/5870/5970/GTX295 and that's it , GTX480 isn't even stores and i dont think it will catch up to anything , so if Ati has more market share in the area where more perf. is available so details can be activated and more eye candy can be done , why they would favor Nvidia ?
Also , Nvidia doesn't give free help , only stupid developers can understand this as free , they come "helping" turning an easy on hardware game into a resource hungry game that needs a better graphic card from costumers and this makes sales bad for the developer , also they force them to not cooperate with Ati , don't give me statemens from Nvidia now , they are so credibile like the charts they give to the press , again the dev. loses potential buyers , look at metro 2033 requirements
Recommended:

Any Quad Core or 3.0+ GHz Dual Core CPU

DirectX 10 compliant graphics card (GeForce GTX 260 and above)

2GB RAM

and the optimum
Optimum:

Core i7 CPU

NVIDIA DirectX 11 compliant graphics card (GeForce GTX 480 and 470)

As much RAM as possible (8GB+)

Fast HDD or SSD

Do you think these people will make money after this game ? do you think the fellow with 8800 will go buy this game ? or the fellow with the 9600GT ? even the one with GTX260 will have a hard time convincing himself to buy this game after what he read on the internet.
It's a game that shines only when you activate all the eye candy , sets the mood better , it's not a great gameplay game like COD MW2 or BC2.
Not all Nvidia deals can be so hurtfull to costumers who have Ati cards , some games accept Nvidia's help on their terms like help us and will put your logo there and that's it , those are few and can deal like this because they have a good game on their hands.

I'm sorry, but you are still completely incorrect. It is free help. The devs that use Physx, already decided to do so ahead of time, therefore nV's help is just a bonus.

Nv also helps on titles that don't support Physx.

And yes, devs will still favor nVidia because of market share. Advanced features can easily be disabled if your system can't use them.

Again, the number of people buying games on lower end hardware proves you wrong. There are far more casual gamers with mid to low hardware than top end. The game devs will ALWAYS factor in these systems.

Metro 2033 can be run at lower settings, and it will be run at lower settings for those that want the game but don't have top end hardware. Which is most of them, btw.

Again, you are quite simply incorrect here.

Did you read the thread title by chance? Just wondering.


But with the logic you are showing, DX11 doesn't matter as Nvidia still pwns with total number of cards in the world so I will buy a Fermi as the 8800, er 9800, er whatever from Nvidia is great!!!!!


So lets never move ahead, lets never make anything new and better.


Is Fermi a game changer? No, it doesn't implement anything new, different, or with a large enough difference between the competition to make jack all of difference. 32AA is marketing gimmic, their ideas about their hardware and implementation have failed. At the uber high resolution people that have these play at Fermi is nothing more than a overpriced hairdryer.

Never said Fermi is a game changer, but neither is the 5k series yet. DX11 still doesn't have significant market share for devs to ignore DX9 and DX10 hardware is my only point.

I was only responding to Leonard's post about how game devs will ignore nVidia now, which is quite simply not true.

 

[erocker]

We do not need people arguing back and forth with an argument that won't ever end. With a poll like this it is best to make your vote, optionally state an opinion and move along. With a thread with somewhat hostile posts, it makes other members weary of posting their thoughts.

This would be appreciated from everyone who has already posted in this thread.

Thank you.

 

[Dunceiam]

At this point, I'm arguing against a brick wall. Your entire argument is based on falsified 'facts' that you wholeheartedly believe to be true, and your love for ATI biases everything that comes out of your mouth.

I honestly believe we could go on arguing for pages, so I'll spare everyone that and just end it here.

I bid you adieu.

 

[LAN_deRf_HA]

I suddenly feel motivated to break it down as factually as possible.

5870 vs 480:
AA - 480 wins. Low performance hit, slightly better quality.
AF - 5870 wins. For not being angular dependent.
Heat - 5870 wins.
Power usage - 5870 wins.
Noise - 5870 Wins.
Memory size - 480 wins. For having a higher default size (good for us GTA players).
Noise - 5870 wins.
Drivers - Counting the past few years, I'd say Nvidia. Looking at the here and now, I'd say tie.
Performance - 480 wins. 10% lead overall.
Price - 5870 wins.
Overclock - 5870 wins. I looked at many reviews to see the max percentage of overclock for the 5870 and 480. Then I compared the percentage of overclock to the percentage of gained performance. On average the 5870 showed a larger performance gain from overclocking by about 3%.
Features - Draw. Both cards have a lot of features, with different specialties and some overlap. This is far too dependent on the needs of the individual user.

While this doesn't look good with fermi only winning 3 out of 10 (not counting the draws), what it comes down to is how important are those 3 to you personally.

 

[theorw]

Here the cheapest 5870 costs 342EUR while the cheapest 480 costs 508EUR.
Does the performance difference justify 166EUR ???
U get a PH x4 with that money!

 

[copenhagen69]

I think its a game changer ... as in it help ATI retain the crown as king

All the delays and all the misinformation from nvidia along the way and they release something that can barely keep up with the 5870? I expected more from them ...

to top it off their cards, unlike ATI 5 series, are a step backwards in tech ... heat and power guzzling cards? no thank you

 

[pantherx12]

I can't believe Leonard was calling that other chap a fanboy, all but one of his statements applied to both Nvidia and Ati XD

( the AA thing)

I've not seen any reviews comparing 32xaa so I can't really comment on that.

But @ 16x AA there is clearly no difference between ati and nvidia cards, its more dependant on game engine rather than card.

 

[the54thvoid]

Anyone want to help me mop up the pulled out hair and smashed in teeth lying around after that posting spat?

And I'm not a NV fanboi but i am JSH's lovechild (and he called me Fermelita - cos i'm noisy, drain his resources and hey, I'm hot.!)

This post would've been better if I wasn't a straight guy

 

[copenhagen69]

Anyone want to help me mop up the pulled out hair and smashed in teeth lying around after that posting spat?

And I'm not a NV fanboi but i am JSH's lovechild (and he called me Fermelita - cos i'm noisy, drain his resources and hey, I'm hot.!)

This post would've been better if I wasn't a straight guy

no worries your not ... so the post is great

 

[FreedomEclipse]

58 (3.22%) of you voted "Yes, its the better card" - How so i wonder?? Im suprised that option got any votes at all judging by the whole library of reviews that are available on the internets.

you either cant read or just plain stupid IMO. but its your money & its upto you how you want to spend it so gg.

 

[amd/atifiend]

i say nay on the 480. i have 4870 cfx and will probably go 5870 cfx when they get cheaper. pricewise i have never been a nv fan but they have made great hardware in the past and furthermore have great drivers but that rift is almost gone now and AMD is on the right track with their graphics. I'd be curious to see if AMD decides to be dirty and drop the prices anyway and im also curious to see if they are gonna do the cherry picked 5870 derivitaves like they did with 4890.

personally what i hope to see in the near future is a 5890 style card with a native dual core chip to shrink the size of the cards themselves and have that same top level performance.

designwise and pricewise the 480/470 is a fail. driverwise im sure they will be good but that isnt enough for me to justify the price when the components on my mobo are melting from the heat.

 

[newconroer]

Im sorry the person who posted right before me, i couldnt get through a quarter of what he was saying before he bored me to sleep.

Which is why the interwebs itself has become ;yawn;. The attitude of 'omg that guy wrote an essay' because it was something longer than twenty-five words.

If you're not interested, then just don't post at all.


On the topic, yes it will be an actual game changer for me, because I plan to SLI them. Which will be a lot more effective than my current 4870 X2. Additionally, I should -for the most part- be able to take advantage of Physx, DX11 features and etc. without having to cut back on image quality for performance reasons.

I never believed in SLI or Crossfire in the past, scaling has always been an issue. But Nvidia really tackled this aspect of multi-GPU systems, by providing us a product that truly does scale well, and for that, I am going to give them my business.

Make no mistake, Crossfire 5870 has been bouncing around my head a while now, but this is a complete system rebuild, and I'm getting pretty much all the best toys. The lack of Physx and DX11 focus, is not something I want to suffer through(no matter how irrelevant or relevant people think those aspects are now).

Some day, ATi may be on board(no pun...) with a form of physics processing, as well as rework their GPUs to better focus on DX11 features, but until that time I will give the green camp another try.

And why not?

I don't see what all the fighting is about. For single card solutions...:

5970 - Extremely Expensive $600+?
GTX 480 - Very Expensive $499
5870 - Expensive $400+
GTX 470 - Costly $350
5850 - Somewhat costly $300

5970 - Extremely Fast
GTX 480 - Very Fast
5870 - Fast
GTX 470 - Quick
5850 - Somewhat quick

The GTX also provides better DX11 support, Physx, Cuda, and quite simply, a blind excuse for people to return to the green camp - if they so wish just for the sake of it.
But the former attributes are something ATi cannot provide, and really, for $100 more give or take, you're getting a faster card that carries those aspects.

What the hell is such a big deal about $100? This is the same old nonsense; people wanting premium quality and premium products but then complaining about their price.

Technically speaking, the GTX 480 is the best/most advanced card ever made by Nvidia, and it's only $499. OMG $499 for a graphics card ;faints;.
We use to pay $600 for 3DFX cards ten years ago...

Look, either buy it, or shutup.

 

[leonard_222003]

On the topic, yes it will be an actual game changer for me, because I plan to SLI them.

Great man , don't forget to comeback with some pictures or it didn't happened , people worry about these cards just working alone ( hot and all ) , imagine how two of them would look in a case , VERY HOT !!!!

I never believed in SLI or Crossfire in the past, scaling has always been an issue. But Nvidia really tackled this aspect of multi-GPU systems, by providing us a product that truly does scale well, and for that, I am going to give them my business.

Where does it scale well ? in old games that got optimized in years ? take a new game that pops up overnight and you will have to wait for profiles or driver updates , this will never change for multiple GPU's from this generation.

Some day, ATi may be on board(no pun...) with a form of physics processing, as well as rework their GPUs to better focus on DX11 features, but until that time I will give the green camp another try.

rework their gpu's to better focus on DX11 features ? i didn't known there is anything wrong with the Ati cards , they are actually very suited for games , look how much perf. they give per W compared to Nvidia , seems like Nvidia has to rework their GPU's and not Ati , look how hot GTX480 and noisy is compared to Ati , looks like Nvidia has to rework their GPU's and be more efficient , consider GTX480 has 1 billion more tranzistors and give only 10-15 % more perf. compared to HD5870 , what arhitecture is more efficient ? are you crazy man ?

The GTX also provides better DX11 support, Physx, Cuda, and quite simply, a blind excuse for people to return to the green camp - if they so wish just for the sake of it.
But the former attributes are something ATi cannot provide, and really, for $100 more give or take, you're getting a faster card that carries those aspects.

what are you talking GTX provides better DX11 support ? from perf. point of view or what ? HD5970 is faster so with that logic Ati provides better DX11 support , is it tesselation ? from what i've read no game will ever use that complex level of tesselation as unigine heaven , it's not practical for the image improvements and no game will overuse this feature to the point framerate will crawl at 30 fps on a 500$ card and the graphics will look like well unigine heaven , it looks like shit compared to actual games.

What the hell is such a big deal about $100? This is the same old nonsense; people wanting premium quality and premium products but then complaining about their price.

People complain about gas , people spend 100$ on the whole video card , this might come as a shock to you but some people's lives don't revolve around video cards , 100$ might be much for some and usually most of them.

Technically speaking, the GTX 480 is the best/most advanced card ever made by Nvidia, and it's only $499. OMG $499 for a graphics card ;faints;.
We use to pay $600 for 3DFX cards ten years ago...
Look, either buy it, or shutup.

This is typical brainwashed brain by brands , OMG the new apple USB cable , we should all go buy it because it's apple and the best , OMG the new shoe lace from Nike , we should all buy it like crazy people on acid .
It's Nvidia's best card but unfortunally not the world best card , you said yourself HD5970 is the fastest.

 

[cbupdd]

Great man , don't forget to comeback with some pictures or it didn't happened , people worry about these cards just working alone ( hot and all ) , imagine how two of them would look in a case , VERY HOT !!!!

Where does it scale well ? in old games that got optimized in years ? take a new game that pops up overnight and you will have to wait for profiles or driver updates , this will never change for multiple GPU's from this generation.

It's true.. Link
Checked on other reviews, and it sometimes exceeds the double. It's envidiable.
Or maybe there is a conspiracy and sli sucks.

This is typical brainwashed brain by brands , OMG the new apple USB cable , we should all go buy it because it's apple and the best , OMG the new shoe lace from Nike , we should all buy it like crazy people on acid .
It's Nvidia's best card but unfortunally not the world best card , you said yourself HD5970 is the fastest.

LoL

 

[copenhagen69]

58 (3.22%) of you voted "Yes, its the better card" - How so i wonder?? Im suprised that option got any votes at all judging by the whole library of reviews that are available on the internets.

you either cant read or just plain stupid IMO. but its your money & its upto you how you want to spend it so gg.

LMAO ... this makes me laugh ... not because your are wrong in any way but because it is so spot on

 

[lism]

Did nvidia on purpose disabled the extra Rops/shaders instead of 512 having only 448?

Wizzards own review with the maximum board power listed at 250W while it was far exceeding the 300Watts?

I think the big ass chip is certainly a pain in the ass, and the engineers at nvidia basicly tried to make the best out of it even tho in alot of games its not an exact type of killer. Comparing a single GPU vs a Dual-gpu board is'nt a logical step too, AMD decides to create GPU's that consume lesser power, smaller DIE-size and lesser transistors, nvidia starts to create a 3 billion transistor chip with so much pain in the ass it costed over a half year to put the chip in actual public. The 50Mhz core speed trick is just a coverup for the idle-speed. AMD on behalf created a chip that actually idles along saving real power instead of being clocked down all the way to 50mhz.

I might say AMD is the clear winner, nvidia has an attitude for a couple of years now with rebranding, making false gpu's that chrashes laptops, quitted in making chipsets for several platforms because the other party often made better performing chipsets at lower power usage.. i can't blame nvidia for it all but rebranding several chips to be at least back in the OEM-game of for example dell is just a joke.

 

[Wile E]

58 (3.22%) of you voted "Yes, its the better card" - How so i wonder?? Im suprised that option got any votes at all judging by the whole library of reviews that are available on the internets.

you either cant read or just plain stupid IMO. but its your money & its upto you how you want to spend it so gg.

I don't see it as stupid to buy it, depending on your budget, and primary usage. Bet this would crush ATI as a folder, or be good as a GPGPU card. It is also faster in games, and still consumes less power than a 4870x2, which is still a relevant card in the market.

It is entirely too hot and loud tho. And the power consumption could be much better, despite there being worse on the market. I still don't think it's a terrible card, it's just not a great, game changing card.

 

[Binge]

I don't see it as stupid to buy it, depending on your budget, and primary usage. Bet this would crush ATI as a folder, or be good as a GPGPU card. It is also faster in games, and still consumes less power than a 4870x2, which is still a relevant card in the market.

It is entirely too hot and loud tho. And the power consumption could be much better, despite there being worse on the market. I still don't think it's a terrible card, it's just not a great, game changing card.

I have to agree with Wile E even though I won't buy the card myself. This whole enthusiast bit is a scene. I feel the scene changed when ATI released the 5XXX series, and I felt satisfied when a cheap card could OC to eyefinity and beyond while staying cool and quiet. Even feeling that deep in my bones won't get me to bash Wile E for stating facts and backing his enthusiast view with sound logic.

 

[leonard_222003]

It is also faster in games, and still consumes less power than a 4870x2, which is still a relevant card in the market.

Is 4870x2 still done by AMD/Ati ? or there are some remaining cards in stores ? Comparing 55 nm tehnology (4870x2 ) with 40 nm tehnology (GTX480) in power consumption ? isn't this wrong a bit ?

 

[Wile E]

Is 4870x2 still done by AMD/Ati ? or there are some remaining cards in stores ? Comparing 55 nm tehnology (4870x2 ) with 40 nm tehnology (GTX480) in power consumption ? isn't this wrong a bit ?

It's still relevant because it still performs up to par with all but the highest of today's cards.

And the process a card is built on doesn't matter, only the end results matter. In this case, both cards are power hungry as hell. lol.

 

[leonard_222003]

And the process a card is built on doesn't matter, only the end results matter. In this case, both cards are power hungry as hell. lol.

It matters a lot for power consumption , it's not fair to compare these two cards when one is built with older tehnology , i know know you wanna spin the bad to good out of GTX480 how is not that bad , how 4870x2 consumes just as much and bla bla bla , problem is the card isn't made anymore and won't be missed by anyone as long as 5000 series is very good and offers everything the old ones plus more.