[VR] NVIDIA GeForce GTX 680 Specifications Revealed

[badb0y]

If nVidia can cream AMD at the same price points they would. The fact that GK110 isn't ready has nothing to do with AMD so stop trying to shift the blame here. If the GTX 680 is GK104 based then I will wait till the refresh to pick up Kepler.

 

[thilanliyan]

Woof.

Fact #1: Que? You only going on BF3 numbers again? That fact is definitely, not a fact.
Fact #2: I just bought a Sapphire Ref for $560, in a laundry list of cards with only one reaching $600 (Giga Windforce 3) another fact that is not a fact.
Fact #3: 8800 Ultra and GTX 280 all had +$600 MSRPs. Fact #3 is again, not a fact.

Wow guy, 0 for 3. Troll much?

Tactic?

Guess if you keep repeating the same lines over and over you'll believe yourself?

It's something Wreckage used to do. Just keep repeating something and make the statement more and more ridiculous.

And lol at 7970 faster than 580 by only 5%. He must have been talking about the 7950. Did you make a mistake OILFIELD?

 

[blackened23]

Tactic?


It's something Wreckage used to do. Just keep repeating something and make the statement more and more ridiculous.

And lol at 7970 faster than 580 by only 5%. He must have been talking about the 7950. Did you make a mistake OILFIELD?

1 benchmark out of 300 under highly specific conditions makes it true. Also, AMD caused the delay of GK110. Undercover agents working at nvidia and such bros. Currently secret plans to manipulate kepler pricing by said AMD agents working at nvidia are also underway.

 

[lavaheadache]

Fact: 7970 is around 5% faster than a gtx 580 at 1920X1200
Fact: 7970 is a 50% increase in msrp and most are priced even worse at nearly $600
Fact: 7970 is the highest msrp since the gtx 280 debacle which nvidia offered rebate checks because it was priced so badly. I'd say even a worse value than the $800 8800gtx. At least the 8800 was a beast in its time. The 7970 is a minor improvement over a 15 month old 40nm product.

/Dwight Schrute

I've been steering clear of these threads for a couple weeks because I've grown tired of reading this stuff. Not only has it been repeated a million times but it is so wrong I don't even know where to begin.

At no point of ownership of my 7970 did I notice it only being 5% faster than my previous 580. My 580 was great and they still are very good cards but the 7970 is clearly in a different league esp. overclocked.

I'm half tempted to mail you my card for a week so you can get all this nonsense out of your head.

 

[railven]

I'm half tempted to mail you my card for a week so you can get all this nonsense out of your head.

Don't do it, it's a trap!

Well makes sense, talk trash in a forum, get a loaner card, magically "Guess I was wrong!" Yet user gets a free upgrade.

:hmm: Now to implement this plan that is so crazy it just might work. :hmm:

 

[BallaTheFeared]

1 benchmark out of 300 under highly specific conditions makes it true. Also, AMD caused the delay of GK110. Undercover agents working at nvidia and such bros. Currently secret plans to manipulate kepler pricing by said AMD agents working at nvidia are also underway.

13% according to one of the more comprehensive test suites out there, which was limited to DX11... I could see that dropping down a bit with 7970's lack luster DX9 performance.

 

[railven]

13% according to one of the more comprehensive test suites out there, which was limited to DX11... I could see that dropping down a bit with 7970's lack luster DX9 performance.

Missed the mark by 8%.

Sorry, you lose. you get nothing. Good day sir!

EDIT: Oh, remind me to actually look at the graphic next time haha. I can no read

You need to check your math, and I thought mine was bad

 

[blackened23]

OH BOY

This is a trap

We should start a back and forth with me getting 20 benchmarks showing the 7970 destroying the GTX 580 by 25% at stock clocks in both DX 9 / 11 games, then you retort with 10 benchmarks showing 5% speed boosts in other games, and then we go back to a price performance argument, then we'll start comparing GTX 580 / 7970 to 470 SLI overclocked to 900mhz.

The circle is complete

 

[blackened23]

Missed the mark by 8%.

Sorry, you lose. you get nothing. Good day sir!

Techpowerup, where the GTX 570 outperforms a GTX 590, and they also benchmark the 7970 using December 2011 drivers that don't support xfire in skyrim. Good stuff!

We should bust out the 1280 x 720 benchmarks while using OLD drivers techpowerup style. Techpowerup testing methods are clearly the best in the biz :thumbsup: Its the ONLY way to benchmark :thumbsup::thumbsup:

 

[Quantos]

Techpowerup, where the GTX 570 outperforms a GTX 590, and they also benchmark the 7970 using December 2011 drivers that don't support xfire in skyrim. Good stuff!

We should bust out the 1280 x 720 benchmarks techpowerup style. Using old drivers at that, techpowerup testing methods are clearly the best in the biz : thumbsup: Its the ONLY way to benchmark :thumbsup::thumbsup:

Wait a minute sir, not all 720p benchmarks are valid. Only SCII provides the most accurate information. SCII at 720p should be the one and only benchmark that decides how everybody buys their card.

 

[BallaTheFeared]

I'm sure you can find some games where it's 25%, just like I can find some where it's slower by nearly as much.

 

[railven]

Techpowerup, where the GTX 570 outperforms a GTX 590, and they also benchmark the 7970 using December 2011 drivers that don't support xfire in skyrim. Good stuff!

We should bust out the 1280 x 720 benchmarks while using OLD drivers techpowerup style. Techpowerup testing methods are clearly the best in the biz :thumbsup: Its the ONLY way to benchmark :thumbsup::thumbsup:

I'd rather not get into a pissing match about reviewing sites. Frankly, anyone will use anything just to prove a point, regardless if the point is contradicted by a separate review in the same review article originally used.

I just learn to ignore those kind of posts, since they're only trying to prove something and will stop at nothing to try to prove it.

I stick to AT reviews, since I trust them, and it doesn't introduce variables when trying to compare to another source.

 

[railven]

I'm sure you can find some games where it's 25%, just like I can find some where it's slower by nearly as much.

Challenge Accepted!

And in the one you linked it was <2% slower. Just saying.

EDIT:

I WIN THE INTERNET!!!

 

[blackened23]

I'd rather not get into a pissing match about reviewing sites. Frankly, anyone will use anything just to prove a point, regardless if the point is contradicted by a separate review in the same review article originally used.

I just learn to ignore those kind of posts, since they're only trying to prove something and will stop at nothing to try to prove it.

I stick to AT reviews, since I trust them, and it doesn't introduce variables when trying to compare to another source.

Well, i'm not trying to stir the pot (much ) but really, why would anyone care about low resolution benchmarks? 1024 and 720 matters in 2001 but its largely irrelevant now.

Also quoting TPU benchmarks when they're using old drivers for both brands that don't support SLI and xfire in skyrim seems silly. At least AT tries to properly test GPU's, TPU has unconventional testing methods to say the least. And not using the latest drivers for both brands is pretty unexcusable.

 

[railven]

Well, i'm not trying to stir the pot (much ) but really, why would anyone care about low resolution benchmarks? 1024 and 720 matters in 2001 but its largely irrelevant now.

Also quoting TPU benchmarks when they're using old drivers for both brands that don't support SLI and xfire in skyrim seems silly. At least AT tries to properly test GPU's, TPU has unconventional testing methods to say the least. And not using the latest drivers for both brands is pretty unexcusable.

Understandable, and I can see where you're coming from. Frankly put, people just link to wherever they want, like you said - they'll find the 1 benchmark out of 300 that proves their point.

Link it, then smile like they just won the internet. Hey, I just did that

EDIT: I hit post and then saw my post before yours. Holy Space-Time-Continium Batman!

 

[BallaTheFeared]

AMD finally got around to adding CF support to Skyrim?

Do you have a review where it's showing the performance increase CF brings with AO enabled?

 

[railven]

AMD finally got around to adding CF support to Skyrim?

Do you have a review where it's showing the performance increase CF brings with AO enabled?

Goal posts, officially moved.

 

[blackened23]

AMD finally got around to adding CF support to Skyrim?

Do you have a review where it's showing the performance increase CF brings with AO enabled?

Xfire has worked in skyrim for some time now, yes. I get 90%+ gpu usage on both GPU's using the latest CAP. I think AMD improved in this respect quite a bit, it was a big issue with cayman cards. ME3 had day 0 support, as did Alan Wake (nvidia still doesn't have an Alan Wake profile btw )

 

[BallaTheFeared]

So between 12/11 and 3/12 they added CF support? Which CAP was that, and how does AO look in Skyrim on AMD cards?

 

[railven]

So between 12/11 and 3/12 they added CF support? Which CAP was that, and how does AO look in Skyrim on AMD cards?

Definitely not as good as it looks in BF3.

 

[BallaTheFeared]

I wouldn't know anymore, I had to disable it

 

[blackened23]

So between 12/11 and 3/12 they added CF support? Which CAP was that, and how does AO look in Skyrim on AMD cards?

I imagine AO looks just fine since AMD cards handle AO just fine. Additionally, skyrim crossfire support has been there since 11.11c but the initial CD driver for the 7970 disabled it. I'm not sure which release but I believe rc11 fixed that.

Google catalyst creator, he mentioned skyrim xfire support as of 11.11c.

 

[BallaTheFeared]

I was hoping to see a review, are there any that show the scaling?

AMD's been on the move, they got CF to work and added driver like AO for skyrim. That's pretty impressive.

 

[lavaheadache]

Xfire has worked in skyrim for some time now, yes. I get 90%+ gpu usage on both GPU's using the latest CAP. I think AMD improved in this respect quite a bit, it was a big issue with cayman cards. ME3 had day 0 support, as did Alan Wake (nvidia still doesn't have an Alan Wake profile btw )

Just wanted to chime in and say that Alan Wake is surprisingly a beast of a game. Even my POS 7970 could use a little more umph @ 2560x1600 with 8x aa. I'm sure my definition of needing more umph is very different from a lot of peoples though. I was getting as low as mid 40fps in heavy areas of the game.

 

[lavaheadache]

Challenge Accepted!

And in the one you linked it was <2% slower. Just saying.

EDIT:

I WIN THE INTERNET!!!

Not that I want to derail your point but those benchmarks don't include Physx, needless to say a 580 or 7970 would need an add in Physx card for optimal performance at high res anyway with Physx enabled.