I can remember the nVidia proponents not accepting O/C'ing back then because it was only guaranteed performance that mattered.
You sir, clearly have an issue at understanding English because that's not what I said.
Now, is this a character assassination because you are claiming I said those things, which is totally not true.
Again, there's gotta be rules you are breaking here right? Am i allowed to go around paraphrasing other members and making up total bullcrap and claiming its what they said?!
For the record, it's totally unfair to pit OC vs stock, because stuff like this:
Yeah that's a >40% performance gain from OC!
And this, from the cheap 7950 OC wrecking 7970 (which was on-par with the 680) massively.
http://www.hardocp.com/article/2012...lack_edition_video_card_review/8#.Vt0xfvl96Uk
Good review sites should keep their testing to stock v stock, OC v OC. I hope you understand now why that approach is the standard for tech journalism.
You sir, clearly have an issue at understanding English because that's not what I said.
Now, is this a character assassination because you are claiming I said those things, which is totally not true.
Again, there's gotta be rules you are breaking here right? Am i allowed to go around paraphrasing other members and making up total bullcrap and claiming its what they said?!
For the record, it's totally unfair to pit OC vs stock, because stuff like this:
Yeah that's a >40% performance gain from OC!
And this, from the cheap 7950 OC wrecking 7970 (which was on-par with the 680) massively.
http://www.hardocp.com/article/2012...lack_edition_video_card_review/8#.Vt0xfvl96Uk
Good review sites should keep their testing to stock v stock, OC v OC. I hope you understand now why that approach is the standard for tech journalism.
I can remember the nVidia proponents not accepting O/C'ing back then because it was only guaranteed performance that mattered.
Some of the same individuals even went so far as to find a negative spin on the insane OC headroom on those AMD GPUs.
Remember stuff like "AMD didn't clock it high enough" or "AMD just too conservative with clocks"... like it's so bad that these GPUs can do 40-50% OC. -_-
Now that AMD clocks their recent stuff higher, it's all like "can't OC at all" or "no headroom".. which isn't even true, Hawaii & Fiji reaches 1.2ghz or above with added voltage. That's basically ~15% to 20%+ headroom right there.
Show me any benchmark of Fury reaching 1.2ghz or this is another of your false claim?
Show me any benchmark of Fury reaching 1.2ghz or this is another of your false claim?
Some of the same individuals even went so far as to find a negative spin on the insane OC headroom on those AMD GPUs.
Remember stuff like "AMD didn't clock it high enough" or "AMD just too conservative with clocks"... like it's so bad that these GPUs can do 40-50% OC. -_-
When I make a claim* (which I hope you understand, in English, is different to saying "I hope" or "it could be" or "maybe", which are speculative and not a solid statement of fact), it's backed by evidence.
1215mhz (1.215ghz) core and 560mhz vram. Good GCN overclocking needs vcore access.
Hopefully you have a change of mind and understand why review sites that love to use OC vs stock GPUs is simply bad tech journalism.
A reminder, in-case you don't get it (~>40% performance gain vs stock!):
*You are free to speculate or opinionate as you have, but your wording tend to be claims, without factual evidence to back you up, it makes you seem like a liar. I hope that clears up any misunderstanding between us. Peace.
I am talking about proper benchmark. There is no proper benchmark of fury or fury X reaching 1.2ghz.
I can post for you 100s of proper benchmark of GTX 980 Ti reaching 1500mhz which is a fact and reality. I know you have personal feelings for AMD but posting false assumptions or claim wont help AMD a bit.
Did you try to read it properly.That image was taken from TPU review:
http://www.techpowerup.com/reviews/AMD/R9_Fury_X_Overvoltage/1.html
So Techpowerup doing the testing is not proper benchmark...
But lets just say it, it felt mighty good when you had 7970 with over 1.3ghz clocks (1340mhz in my case) and it just roflstomped every 680 (OC'd or not) back in the days. :biggrin:
No it didn't. People here seems to have pretty bad memory, a max oced lightning 680 was the faster card, the [H] review is there for you.
GPU Boost is guaranteed to hit 1058MHz in most games. Typically, the GPU will be going much higher. We experienced clock speeds in demo sessions that would raise to 1.150GHz and even 1.2GHz in such games as Battlefield 3.
http://hardocp.com/article/2012/07/23/msi_geforce_gtx_680_lightning_video_card_review/9#.Vt1d5vl96Uk
Given their small selection of games, the OC 7970 looks to be doing alright.
In a larger datasets like TPU where the 7970 vs 680 at stock clocks were already close, the bigger OC % on the 7970 would give it an edge across many titles.
Afterall, we're talking 900mhz to 1.3ghz, ~44%, which resulted in massive performance gains:
What's the % OC for the 680? I know the Lightning model with its modified bios was able to clock up to ~1.36ghz boost, above other models.
http://hardocp.com/article/2012/03/...force_gtx_680_video_card_review/#.Vt1gb_l96Uk
They mention even the reference card was getting boost clocks up to 1.2ghz in games.
So what kind of performance % gains do you expect from overclocking the 680, 30%? So no, it ain't faster than an OC 7970 across many games. -_-
Okay, LN2 modified bios, nearly 1.4ghz clocks which was well beyond most 680.
There's like 4 games tested, Battlefield 3, Max Payne, Batman, Witcher 2.
Importantly, one must compare the performance gained compared to the 680 with this OC, because a more accurate result is comparing over many games.
Starting with similar performance at stock, what kind of OC % and OC % performance gains does the modified bios 680 get vs the 7970?
Sadly, [H] does not have the data for this. But assuming near perfect performance scaling, going from 1.2ghz boost as [H] observes for the stock 680, to 1.39ghz is what %? Less than 20%. -_-
Let's be generous and assume the stock 680 only boost to 1.1ghz (very low), an OC to 1.39ghz is ~26%.
A max potential gain of 26% over stock with this LN2 bios 680.
Well below the ~45% OC the 7970 gets.
do not explain logic and facts to these some poster.They compared max oced 680 vs a max oced 7970 and 680 came out ahead.You just can't compare clocks for 680 , memory oc benefited Kepler very well.
They compared max oced 680 vs a max oced 7970 and 680 came out ahead.You just can't compare clocks for 680 , memory oc benefited Kepler very well.
In 4 games. And it was very close in some of them, minus Batman. -_-
You know how easily that could skew the data?
This is why it's always been important to judge GPUs in a wide list of games.
How much performance % gains can the 680 get from OC compared to it's stock performance? Cos the 7970 is quite close to the 680.
If you don't agree that you need a larger sample size than 4 games, think about the current situation, how it would look if a review site took 4 games, picking from AMD favored titles like Hitman, Ashes, The Division, Far Cry Primal, Black Ops 3, Rainbow Six, SoM, Ryse... the result would obviously stack in one side and be completely biased.
Certainly if you agree it's unfair to pit OC vs stock, you would also agree that larger sample sizes are more of an accurate benchmark.
TPU didn't do the Lightning review as [H] did iirc, your chart is true for most of the 680s but lightning was a special snowflake. It required a custom bios to unlock it's full potential but sadly MSI discontinued that later.
Regarding larger sample size for sure but tbh very few sites has the caliber to test games like [H] do so quality over quantity. G3D testes a bunch of gamesbut uses fxaa w/e possible sorry not into that kind of review.
TPU did actually test the Lightning (link). However TPU only tests OC performance in a single game (BF3 in this case), so it doesn't really help determine which card is better across a wider range of games.
680 vs 7970, that takes me back, remembers fondly the wars...
Read the [H] review, TPU didn't use the LN2 bios.