7850 to 660Ti - Weird situation

[SkodaMart]

www.imgur.com

Thanks, uploaded

 

[humbug]

Did someone say my name?

I thought i heard someone say my name....

 

[spixelspixel]

errr... no it doesn't

Heaven:
7950: 57.7
660Ti: 55.3

you're correct about Skyrim at 1200p though (going by my link). Overall, the 7950 wins if you were to add up all the FPS' and compare that to the 660Ti. It's even better at higher resolutions, it can be argued that they are equal at 1080p.

Those heaven results on anandtech have to be wrong. They give 68 fps to the 670 which is what my stock 670 got @ ocuk settings. My 660ti gets 62fps (factory oc, would be 60 at stock) They either weren't tested with the same settings or they're false numbers. I used the earliest drivers on both cards and no difference on newer ones. Since the heaven result is so wrong, its hard to believe any of their other tests as well.

 

[Tonester0011]

Those heaven results on anandtech have to be wrong. They give 68 fps to the 670 which is what my stock 670 got @ ocuk settings. My 660ti gets 62fps (factory oc, would be 60 at stock) They either weren't tested with the same settings or they're false numbers. I used the earliest drivers on both cards and no difference on newer ones. Since the heaven result is so wrong, its hard to believe any of their other tests as well.

Does it say what version of heaven they're using?
3.0 has higher fps at same settings compared with 2.5...etc.

 

[humbug]

Those heaven results on anandtech have to be wrong. They give 68 fps to the 670 which is what my stock 670 got @ ocuk settings. My 660ti gets 62fps (factory oc, would be 60 at stock) They either weren't tested with the same settings or they're false numbers. I used the earliest drivers on both cards and no difference on newer ones. Since the heaven result is so wrong, its hard to believe any of their other tests as well.

yeah, 55 FPS for the 7950 also looks odd, i get that on my FOC 7870.

overclocked... well....

edit, actually it depends on what settings they use.

 

[spixelspixel]

yeah, 55 FPS for the 7950 also looks odd, i get that on my FOC 7870.

overclocked... well....

edit, actually it depends on what settings they use.

Here's my results on cards I've had recently. All 1920x1080 4AA

6870 (stock) - 33fps
560 ti (factory oc) - 41fps (38 stock) -
both of these match anandtechs results also, dont know why their 660ti result is 5fps lower than it should be.

570 (factory oc) - 49fps (46 stock)
580 (factory oc) - 55fps (53 stock)
660 ti (factory oc) - 62fps (60 stock)
660 ti (oc, not to max but near) - 70fps
670 (stock) - 68fps (matches anandtechs result perfectly so I guess they use ocuk settings)
670 (oc, not to max but near) - 78fps

 

[tommybhoy]

Yes, I too could have taken some lovely screen shots, it proves nothing.

The point was to prove that although the card was gubbed, the games still actually played nut performance was all over the place, for example going from 60fps-15fps-60fps in a stuttery fashion, not because of any other fault than a gubbed card.

It wasn't a driver/ hardware design fault, it was because the card was GUBBED.

I tried ATi at my expense and regret it.
When I bought the card I specifically asked if I would have problems with the games I am currently playing, The Witcher 2 and WoT.
I did have problems.

Read above again, bang your head on the desk a couple of times, and eventually the word GUBBED will sink in.

I tried ATi and for me it wasn't as good as nVidia. It is as simple as that and those are my true colours!

We all know your true colours.

I will wait now until the prices of the 660Ti fall significantly then put a 2nd one in.
Unless you game at high resolution (which I don't) 1080p here.
The 660ti is the best bang for buck card currently available, even though I feel it is currently overpriced.

Spoken like a true human being that eats, sleeps, Nvidia.

I will go first then.
Totally stock using the OCUK default settings at 1080p.

His 7950 IceQ@£240+4 free games-which came with the GHz bios@stock:

Note that your mins are less than half of a 7950, get used to it when you apply IQ in game.

Apologies go out if I'm being a bit strong on my words but, it's a sad day when fellow Nvidia users(that really do know there stuff in comparison) are telling you that a 7950's faster and you are still arguing with them.

 

[spixelspixel]

Note that your mins are less than half of a 7950, get used to it when you apply IQ in game.

Apologies go out if I'm being a bit strong on my words but, it's a sad day when fellow Nvidia users(that really do know there stuff in comparison) are telling you that a 7950's faster and you are still arguing with them.

The mins on heaven mean absolutely nothing, if he ran it again it would be completely different, you might have left heaven run a couple of times before starting the benchmark Another time, my mins were 26, and 34. I never let it loop, just instastart

 

[Tonester0011]

Those heaven results on anandtech have to be wrong. They give 68 fps to the 670 which is what my stock 670 got @ ocuk settings. My 660ti gets 62fps (factory oc, would be 60 at stock) They either weren't tested with the same settings or they're false numbers. I used the earliest drivers on both cards and no difference on newer ones. Since the heaven result is so wrong, its hard to believe any of their other tests as well.

Like i've said before, without knowing which version of Heaven they're using, comparing your results with theirs is pointless.

To better show this.

Spoiler

[/URL[URL=http://imageshack.us/photo/my-images/694/00009n.jpg/]

Same settings, same clocks....totally different FPS!
They use the same settings and same version for continuity, that's the scores they got, the 660ti is slower than the 7950.

 

[tommybhoy]

The mins on heaven mean absolutely nothing, if he ran it again it would be completely different,

I have a load of different results from my 6950>70 using different drivers, after just going through them, the mins all sit very close together, so it means something on AMD, but if it holds differently with Nvidia then fair enough, I have just learned something new.

In my defence, I always new the 7950>660ti, add IQ and it just goes downhill

 

[humbug]

@ tommybhoy, get that 7950 overclocked, unfortunately my 7870 is not a good clocker, i can't get it over 1220 Mhz in Heaven, which for a Gigabyte 7870 is pretty poor, they normally clock to 1275+, some as high as 1350, so it's not a good showing

 

[spixelspixel]

Like i've said before, without knowing which version of Heaven they're using, comparing your results with theirs is pointless.

To better show this.

Spoiler

[/URL[URL=http://imageshack.us/photo/my-images/694/00009n.jpg/]

Same settings, same clocks....totally different FPS!
They use the same settings and same version for continuity, that's the scores they got, the 660ti is slower than the 7950.

I know there is differences in scores between versions but my 6870, 560 ti and 670 scores match their results perfectly, using heaven 3.0 @ ocuk settings 1920x1080... so the chances are its what they use.

 

[tommybhoy]

@ tommybhoy, get that 7950 overclocked, unfortunately my 7870 is not a good clocker, i can't get it over 1220 Mhz in Heaven, which for a Gigabyte 7870 is pretty poor, they normally clock to 1275+ some as high as 1350, so it's not a good showing

Can't, it went back because It could be heard in my rig@idle and I'm anal about noise.

I have a 'lazy' 7970 that goes to 1150MHz now, but the 7950 matched it in performance when it was clocked, wish I new at the time an AC extreme fitted it.

 

[spixelspixel]

I have a load of different results from my 6950>70 using different drivers, after just going through them, the mins all sit very close together, so it means something on AMD, but if it holds differently with Nvidia then fair enough, I have just learned something new.

In my defence, I always new the 7950>660ti, add IQ and it just goes downhill

What about the 7950 result on the previous page which shows a min of 36, and its overclocked higher.

Btw, check my max fps, shows the raw power of the 660 ti compared to the 7950 jk

 

[humbug]

Can't, it went back because It could be heard in my rig@idle and I'm anal about noise.

I have a 'lazy' 7970 that goes to 1150MHz now, but the 7950 matched it in performance when it was clocked, wish I new at the time an AC extreme fitted it.

Well that sux you and me both have lazy cards then.

 

[tommybhoy]

I'm on payback for my mental 1GHz+ 6950>70 CrossFires, you win some you lose some.

Feel safe in the knowledge that you won't even notice in game tbph, I'm down 50Mhz your down what 75Mhz?

It's nothing really.

 

[tommybhoy]

What about the 7950 result on the previous page which shows a min of 36, and its overclocked higher.

I don't know how that rig is setup to explain the difference, could be an unstable oc, could be something else, as I said, my min results on file are within ~5fps.

Btw, check my max fps, shows the raw power of the 660 ti compared to the 7950 jk

660 ti (factory oc) - 62fps (60 stock)
660 ti (oc, not to max but near) - 70fps

7950@stock(925MHz)-70fps

As I said, gutted it went back as it performed the same as my oc'ed 7970.

 

[Tonester0011]

I know there is differences in scores between versions but my 6870, 560 ti and 670 scores match their results perfectly, using heaven 3.0 @ ocuk settings 1920x1080... so the chances are its what they use.

Doubt they're using the newest version of Heaven despite your scores lining up.
Best guess would be HWbot Heaven. That's the best Heaven to use if you care for continuity.

 

[spixelspixel]

I don't know how that rig is setup to explain the difference, could be an unstable oc, could be something else, as I said, my min results on file are within ~5fps.





Hows your 660ti oc to near max@70fps cool in comparison to the 7950@stock(925MHz) with 70fps?

As I said, gutted it went back as it performed the same as my oc'ed 7970.

I was being sarcastic about my max fps being better since you were slating skodamart about his minimums. Btw, a 7950 with ghz bios isn't true stock vs stock. Can we get a real 7950 stock result ?

I don't really care about comparisons between the 7950 + 660ti though tbh, just didn't like your reply to skodamart. People always defend whatever card they currently own, nothing new. Both cards are in the same ballpark, that's what matters.

 

[spixelspixel]

Doubt they're using the newest version of Heaven despite your scores lining up.
Best guess would be HWbot Heaven. That's the best Heaven to use if you care for continuity.

Is there a reason why the 670 would be 13 percent faster in heaven 3.0 compared to 22 percent in the hwbot version? Doesn't make sense to me.