GTX480 / GTX470 Reviews and Discussion

[snakej]

well, I would say is wise to wait for the cards itself, then we can see what they will implement as over clocking tools.

let me google that for you

 

[Guest]

There isnt yet but the fact that Heat should be a problem when overvolted on Air,I highly doubt there will be one. Again on water 470 is the way to go if* overvolt programs appear.

 

[OvrClkr]

Lets keep it clean guys, if anyone is in disagreement with a certain post you can always send them a PM. Off-topic/flaming posts will be deleted.

 

[Guest]

Lets keep it clean guys, if anyone is in disagreement with a certain post you can always send them a PM. Off-topic/flaming posts will be deleted.

😀 All I said was proofs lmao not even 1 bad word, just reread his posts 😉 . And no, I wont hide the reality to readers.

He just cant admit that in a certain position, the 5850 is better than a 470.

Where was I off topic lol? Comparing Fermis OC ability on Air.

 

[notty22]

😀 All I said was proofs lmao not even 1 bad word, just reread his posts 😉 . And no, I wont hide the reality to readers.

He just cant admit that in a certain position, the 5850 is better than a 470.

Where was I off topic lol? Comparing Fermis OC ability on Air.

Do you have a guilty conscious ? Its just as brutal to start with , "i'm not talking about another planet', " i'm not talking about water"
'Something smells fishy' when presented with GTX 470 , 480 info.
For all we know, your computer is grey screening, locking up, sporadic rebooting- from you attitude you probably wouldn't even care.
edit : last post, till I get some solid info, I'm feeding a troll here.

 

[Guest]

Do you have a guilty conscious ? Its just as brutal to start with , "i'm not talking about another planet', " i'm not talking about water"
'Something smells fishy' when presented with GTX 470 , 480 info.
For all we know, your computer is gray screening, locking up, sporadic rebooting- from you attitude you probably wouldn't even care.

lol? @ 1Ghz 100% Stable the pics were not enough to prove?, then what does it take to prove you that ? . The grey issues where present with my old 5970 that stopped at Cat 10.2 and was only happening 2times/Week even at default clocks of 725. That was with Cat 9.11 with the grey issue. Never had a single line since I got the 5850 and Cat 10.2. Now, when I say other planet, that means to not hoping in the future about having an over volt tool and to see what we have Right Now. Your post proved me that you did not read what I wrote, When I said fishy about the graphs as they are showing 5870CFX drawing LESS power than a 470 which is wrong, 5870CFX should use more and the 5970 was not included in the test. Thats why I said Fishy.

Sure, notty is not accepting the fact that a 5850 OC better on Air than a 470, I provide him proofs and pics and he STILL refuse the Reality and call me a troll.

Anyway, Any custom cooler for the 480 ?

If you provide me with water cooling for free, I will buy the 480 😀

Edit: I hope this will further prove that a 5850 do 1Ghz stable. Metro2033 http://www.youtube.com/watch?v=HOLjROF_o0g Just Cause 2 http://www.youtube.com/watch?v=76sMHMhzzZU

 

[OvrClkr]

😀 All I said was proofs lmao not even 1 bad word, just reread his posts 😉 . And no, I wont hide the reality to readers.

He just cant admit that in a certain position, the 5850 is better than a 470.

Where was I off topic lol? Comparing Fermis OC ability on Air.

I didn't give out any names elie, my point is in "general" 😉

 

[RealityRush]

elie, go read some user reviews of the 480/470, they are OCing on air just fine.

A 480 (not sure about the 470) could overvolt and reach 1GHz on air, it's just you'd be running the fan at 100% and be in the high end of the 90s for temps. But the card wont die.

Regardless of that fact though, even if a 470 can only get to 850 stable without the fan being loud and annoying (which I already linked you proof that it is possible once an overvolting program is released) a 200-250MHz increase on a 400 series card might actually give you a bigger performance boost than 300-350MHz on a 5800 series card due to the fact that the 400's do more work per clock.

Your 5850 may OC higher, but an OC'd 470 on AIR is probably still going to beat it.

I'll let you know when I get mine how it goes, then we can all stop arguing emotions rather than hard facts, hopefully there is an overvolting tool by the time I get my cards. If I brick them, well, I'll figure out how to get the lifetime warranty to cover it 😛, I'm crafty that way. I'll even ramp up all my case fans and GTX 470 fans (both of em) and get em to 1GHz in SLI and screenshot it for you 😛. At the end of the day, I realize that means nothing, but seeing your reaction could be fun 😛!

 

[Guest]

elie, go read some user reviews of the 480/470, they are OCing on air just fine.

A 480 (not sure about the 470) could overvolt and reach 1GHz on air, it's just you'd be running the fan at 100% and be in the high end of the 90s for temps. But the card wont die.

Regardless of that fact though, even if a 470 can only get to 850 stable without the fan being loud and annoying (which I already linked you proof that it is possible once an overvolting program is released) a 200-250MHz increase on a 400 series card might actually give you a bigger performance boost than 300-350MHz on a 5800 series card due to the fact that the 400's do more work per clock.

Your 5850 may OC higher, but an OC'd 470 on AIR is probably still going to beat it.

I'll let you know when I get mine how it goes, then we can all stop arguing emotions rather than hard facts, hopefully there is an overvolting tool by the time I get my cards. If I brick them, well, I'll figure out how to get the lifetime warranty to cover it 😛, I'm crafty that way. I'll even ramp up all my case fans and GTX 470 fans (both of em) and get em to 1GHz in SLI and screenshot it for you 😛. At the end of the day, I realize that means nothing, but seeing your reaction could be fun 😛!

😉 I hope you break the 21K barrier. I see you are too confident about getting them to 1Ghz in SLI on Air. Looking forward for your results [:bohleyk:1]

Finally som1 than can give proofs. I think a 470@880Mhz should be around the 5850@1Ghz performance.

 

[RealityRush]

😉 I hope you break the 21K barrier. I see you are too confident about getting them to 1Ghz in SLI on Air. Looking forward for your results [:bohleyk:1]

😛 I'm always supremely confident hehe. When I OC'd my Phenom II x4, I didn't slowly burn that mother in, I just cranked it up on water and went nuts, it is more fun that way, I never know what is going to happen. A super clocked CPU? A BIOS reset? A paper weight? Who knows!?

But yeah, I still think 1GHz is achievable on air considering I'll have 4 120mm fans blowing fresh air directly into my cards 😛. Would it be stable enough for daily use? Probably not.... but that's beside the point, a gauntlet has been thrown down and I feel the need to pick it up . I am considering water cooling them in the future.... but then it occured to me I can just hardcore air OC them with all my airflow in my case and save the money on cooling for trading up to refreshes and then buy aftermarket cooling for those

Finally som1 than can give proofs. I think a 470@880Mhz should be around the 5850@1Ghz performance.

I would agree that the improvement from OCing would likely be about the same at those speeds. And if someone got 800 without overvolting 880 should be easily within reach. Hey, we agreed on something! Wewt! 😛

All that needs to happen now is Direct Canada needs to get some bloody stock so I can actually purchase something... why the heck is everything always more expensive and delayed up here in Canada >.<

 

[shadow187]

Clock-for-clock on the 470 and the 5850, we know who is the winner. Ability to OC on the cards, the 5850 wins. The 5850 can OC to the level which it is equal to an OCd 470. I can't wait to see a review of that.

The only bad part is, I do not see a dual-GPU solution coming out soon.

 

[Guest]

There should be a thread about Fermi refresh preview. If nVidia lowers the TDP,28nm? , they can squeeze more Mhz and enable Full 512 Shader. A 480@512Shaders@850Mhz will beat the 5870 by about 30% and will be near the 5970 performance. nVidia still needs a Dual-GPU to take back the fastest single card crown.

 

[RealityRush]

OC's don't count unless it is stable enought to;

Beat Crysis start to finish, with no puases/restarts.

... some of the biggest OC's in the world on ln2 could be considered unstable and outside of that margin, so they are invalid?

This is why I said there's too many variations of "OC'd" to use OCing to compare cards.

The 470 beats the 5850 at stock... therefore a 470 could beat a 5850 OC'd as well, because however much you OC that 5850 you can OC the 470 as well.

If we are talking strictly air, I already gave you links that show stable OC's of the 400 series cards up to 800MHz at like 83Celsius and the only reason they couldn't go higher was they couldn't voltage tweak their cards... yet. And that was running Furmark, Crysis would be easier still. The 470 doesn't even have to OC as much as a 5850 to match it's % improvement in performance due to more work done per clock.

 

[IzzyCraft]

There should be a thread about Fermi refresh preview. If nVidia lowers the TDP,28nm? , they can squeeze more Mhz and enable Full 512 Shader. A 480@512Shaders@850Mhz will beat the 5870 by about 30% and will be near the 5970 performance.

Well that's the ideal pipe dream you know a good old repeat of the G80 to G92 that gave the 8800gt performance near the 8800ultra for half the price. Lol i'd just be happy if nvidia could make a card that doesn't need to be crippled and sold. ie the whole not having full 512 "cores" out there.

 

[Guest]

Ya we all know clock for clock Fermi is faster but the question is can it be overvolted on air within safe margins, again if an OC is 100% 20 min Furemark stable it will never fail 5850@1Ghz, though I still think that a 5850@1Ghz should be a bit faster than a 470@800, ~880 it should match the 1Ghz 5850. I think that Fermi refresh should answer all those questions.

Edit: The link shows a 480@800@100%Fan not a 470@800@100%Fan...

 

[IzzyCraft]

Yeah i doubt 40nm will be a viable remake just due to one of the issues is that physical size of the chip is so large which is messing with costs and yields. So a die shrink without too much change would do them well although knowing nvidia's history another die shrink seems unlikely for the 1st refresh rather some trimming or something will be done.

 

[builderbobftw]

....That's what you get when you try to jam that many tranistors in a chip and fail to use them correctly....

 

[Guest]

Its like the GTX280->GTX285 transition.

 

[builderbobftw]

Yep.

And the GTX 285 was still trash, Preforms like a 4890 at a boatlod more, gets utterly and totaly OWND by the 4870x2 at 20$ more.

 

[IzzyCraft]

I'd take that the GT200 vs GT200b yeilds and failure rates were vastly different. Nvidia greatly improved upon the yields and slightly the performance of the chip. But it's to note that there was a die shrink.

 

[builderbobftw]

Right, But it was still a vastly inferior product, and methinks history will be reapeting itsself.

 

[OvrClkr]

Yep.

And the GTX 285 was still trash, Preforms like a 4890 at a boatlod more, gets utterly and totaly OWND by the 4870x2 at 20$ more.

Since when does the 4890 compete with the GTX 285?? AFAIK the 4890 is on par with the GTX 275 give or take.

Read the reviews builder 😉

http://www.techreport.com/articles.x/16681/13

 

[shadow187]

There were times when the 4890 hit the 285.

... some of the biggest OC's in the world on ln2 could be considered unstable and outside of that margin, so they are invalid?

This is why I said there's too many variations of "OC'd" to use OCing to compare cards.

The 470 beats the 5850 at stock... therefore a 470 could beat a 5850 OC'd as well, because however much you OC that 5850 you can OC the 470 as well.

If we are talking strictly air, I already gave you links that show stable OC's of the 400 series cards up to 800MHz at like 83Celsius and the only reason they couldn't go higher was they couldn't voltage tweak their cards... yet. And that was running Furmark, Crysis would be easier still. The 470 doesn't even have to OC as much as a 5850 to match it's % improvement in performance due to more work done per clock.

The 5850 can get a higher OC than the 470. Therefore an OC 470 is not equal to an OC HD5850.

 

[RealityRush]

The 5850 can get a higher OC than the 470. Therefore an OC 470 is not equal to an OC HD5850.

Again.... the 470 does more work per clock, so it only needs like 70% of the OC of the 5850 to match its increase in performance. Which it looks like it might be able to, until I get my hands on one or two though we wont know for sure 😛

 

[rawsteel]

607 out of 725 is about 84%

If 5850 clocks to 1100 with voltmods - 470 would have to reach ...... wait for it ...... 924Mhz. Now that is something impossible on air I can guarantee.

Notice this : 480GTX was clocked at 800 on 100% fan reaching 83C. This is only a mild 14% overclock when you consider its stock is 700.

Now how much you think you will get with voltmod 😀. if you just up the voltage with a notch and still run at 800 you will get more heat as more voltage is pumping in, so you will be already in 85C or more.

I really doubt if you will be able to get to 850 even with voltmod, you will just get 100C on 100% fan and I bet you cant game at these settings for like 2-3-4h. It might run tests but try gaming a whole day, you will melt some case metal there

So no, I dont think OC Fermi = OC 5000. ATI have much more headroom.

However I dont think lot of people overclock, let alone so agressive so this probably matters to a very small number of enthusiasts.

What matters is stock vs stock and 470 beats the 5850 hence the price is a little more. Same for 480.