well, from what I've gathered it sucks more power, produces more heat and only beats 5870s by 10-15% ? is it so or am I missing something ?Quote:
Originally Posted by Helloworld_98
Printable View
well, from what I've gathered it sucks more power, produces more heat and only beats 5870s by 10-15% ? is it so or am I missing something ?Quote:
Originally Posted by Helloworld_98
yeah, these little things called drivers and that not everything scales well.Quote:
Originally Posted by rusty
although saying that, the GTX 400 series scales ridiculously well in SLI.
nah you're not missing anything, but while power consumption and heat are important parameters in the "Should I buy this card" function f(x), they are not the only parameters. How much power it takes isn't the only thing that's important about a video card. If you are one of those guys who pay little consideration to how much power a card uses, you might find that the remaining parameters make 480 actually a pretty good buy, performance - feature - price wise.
But if you do find power usage important, which many people do, then the card's attractiveness gets less and less.
5850, 5870 and even 5970 were easily recommendable to anyone with a respective budget, but GF100 looks as if it's more of a personal preference issue.
hmm should I wait till mid summer and get some 480 then or whatever revision it will be then ?
I have made a few diagrams based on ComputerBase.de review to compare GTX285, HD5870 and GTX480 graphics cards.
Test setup
CPU: Intel Core i7 965 Extreme Edition (overclocked via multiplier to 3.87 GHz quad-core)
Motherboard: Asus Rampage 2 Extreme (Intel X58, BIOS-Version: 1639)
Hard Disk: Western Digital Caviar Black 1TB, 32 MB cache
Memory: 3x 2.048 MB Corsair DDR3-1600 (8-8-8-24)
OS: Windows 7 Ultimate 64
PSU Coolermaster M850 Real Power Pro Modular (850 Watt)
URL http://www.computerbase.de/artikel/h...itt_testsystem
Nvidia Drivers: Nvidia GeForce 195.62
GTX 400 Drivers: Nvidia GeForce 197.17
ATI Drivers: ATi Catalyst 9.11 (Dirt-2-Hotfix)
HD5870 ATi Catalyst 10.3a Preview
HD5830 ATi Catalyst 8.703_RC2
HD5670 ATi Catalyst 8.69 RC3
The diagrams are based on average FPS with optimal - in game - settings for different combinations of screen resolution, AA and AF.
The charts have the same size and max scale at 120FPS. This way, the diagrams are comparable, but on the other hand, the 1920x1200 1xAA, 1XAF chart misses the World in Conflict SA 150,3FPS FPS value.
http://i237.photobucket.com/albums/f...0x1050_1_1.png http://i237.photobucket.com/albums/f...x1050_4_16.png http://i237.photobucket.com/albums/f...x1050_8_16.png
http://i237.photobucket.com/albums/f...0x1200_1_1.png http://i237.photobucket.com/albums/f...x1200_4_16.png http://i237.photobucket.com/albums/f...x1200_8_16.png
http://i237.photobucket.com/albums/f...0x1600_1_1.png http://i237.photobucket.com/albums/f...x1600_4_16.png http://i237.photobucket.com/albums/f...x1600_8_16.png
Nice job :up:
Would love to see the same thing done with Donanimhaber review, because I believe they are the only one with OCed HD 5870, GTX 480 and GTX 470.
pretty much what i said higher res + 8aa fermi tanks
http://www.brightsideofnews.com/news....aspx?pageid=0Quote:
Originally Posted by Russian
Quite a decent review, actually. The only thing I'm missing a bit is testing in various resolutions, but that's it.
Yeah, Raven 2 is crap, riiiiight. No wonder I love mine so much, here's the proof! :D :up:Quote:
As you can see, the difference between ATI and nVidia is only seven degrees Celsius. When we loaded the boards, the situation changed radically - in case of 3-Way SLI GeForce GTX 285 and Cooler Master's Cosmos S, the temperature went from 102/98/95C to 84/85/79C. We expected the similar case with GTX 480 and especially GTX 480 SLI. However, the boards worked in nearly identical temperatures in both cases: in Cosmos S, a single GTX 480 would heat up to 101 Celsius, while two boards would result in GPU0 heating to 105C, with GPU1 [lower board] heating to 98C. In Raven RV01, single GTX 480 would heat up to 97C. When both boards were active, GPU0 showed 97C, and GPU1 showed 88C. If you're going with a multi-GPU setup and aren't considering liquid cooling, we seriously recommend getting a Raven RV01 or RV02. In case of buying a complete system, our recommendation would go solely to MAINGEAR and system integrators that utilize Raven RV01 or RV02 chassis
Yea I think that really shows it well. Kind of sucks really if you think about it. For me, it wouldn't matter too much, I'm limited to a 22" screen. But still, it should be the other way around with the 480 . Very strange.Quote:
Originally Posted by eric66
GeForce GTX 480 Gets a BIOS Update and Cooks an Egg
http://www.legitreviews.com/article/1264/1/
I know this is the same old line but...drivers play a role here also, so the numbers may change. Not drastically enough to change the overall consensus though.Quote:
Originally Posted by Tim
http://www.legitreviews.com/images/r...gg_cooking.jpgQuote:
Originally Posted by onethreehill
LMAO! :rofl: :rofl: :rofl:
bit-tech: Fermi Testing Update
http://www.bit-tech.net/hardware/gra...sting-update/1
Good on bit-tech for taking the time to do that. Pretty much validates a bunch of results. :up:Quote:
Originally Posted by onethreehill
That's very contrary to what almost all the reviews say, as the more AA the better Fermi does. However it does have issues with high resolutions (ie. 25x16) but that is supposedly a driver issue.Quote:
Originally Posted by eric66
480 is meh, GTX470 is a much better choice. Better price/performance than a 5870 (and tied with a 5850), not much more power, and scales much better in SLI than Crossfire.Quote:
Originally Posted by rusty
TBH I think the only things that make Fermi worth buying are all of Nvidia's "extra's". PhysX and usually better drivers and all of that.
here cooking an egg on 480 while benching
http://www.youtube.com/watch?v=dQw4w9WgXcQ
'This video contains content from Vevo, who has blocked it in your country on copyright grounds.'Quote:
Originally Posted by rusty
:(
ok try this one thenQuote:
Originally Posted by Helloworld_98
http://www.youtube.com/watch?v=oHg5S...x=0&playnext=1
Re-quoted for emphasis. Been saying this all along....Quote:
Originally Posted by zalbard
it's not nice to rickroll peopleQuote:
Originally Posted by rusty
must be the fan is running quicker for that, to keep temperatures down as the 2 gig version is relatively cool, even compared to the 1gig version :eek:Quote:
Originally Posted by SKYMTL
i suppose if they whacked on the George Foreman grill onto it, they could lower the fan speed if ATI wanted, but the additional expense probably isn't worth it in their eyes.
http://techreport.com/r.x/geforce-gtx-480/gpu-temps.gif
The 2Gb one is a Asus Matrix Radeon HD 5870 2GB with 900 MHz core and 4900 Mhz vram if anyone did not know...
Hm, so are you saying that the reference board is going to operate with less noise? Interesting, possible I guess, although I thought Asus Matrix had a better heatsink.Quote:
Originally Posted by ajaidev
Agreed, it's a nice article to see.Quote:
Originally Posted by SKYMTL
This is from HEXUS's review of 5870 EF6:
"For those that enjoy a quieter computing experience, we found the fan on the Radeon HD 5870 Eyefinity 6 Edition hit speeds of 1,998rpm and remained relatively quiet during a session of Battlefield: Bad Company 2 at 5,040x2,100.
The CrossFire configuration - with two Radeon HD 5870 Eyefinity 6 Edition cards in close proximity - proved to be warmer, resulting in the fans spinning up to a noisy 2,425rpm."
and here is the chart from them in their GTX 470's review:
GeForce GTX 470 2,350rpm/3,290rpm 5/3
GeForce GTX 480 2,720rpm/3,785rpm 4/2
Radeon HD 5970 2,460rpm 5
Radeon HD 5870 TOXIC 1,950rpm 7
Radeon HD 5870 2,350rpm 6
Radeon HD 5850 1,650rpm 7
___________________________________________
Radeon HD 5870 EF6 1,998rpm 7