• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

ATI cuts 6950 allocation

It does sound like AMD are setting themselves up for a bit of a pasting if their new cards underclock themselves by default when they're feeling a bit stressed.
 
Nope, I don't think they'll be underclocked. Looking back at the Perlin Noise results (which is the subject of the AMD slide showing how PowerTune works) from the German forum they did look low compared to the other tests when compared to a 5870. This along with Furmark are likely throttled to maintain the default TDP. However, games rarely reach the power usage of synthetic benchmarks and I doubt AMD would have chosen either a TDP so low, or a clock that was so high which meant that games would be throttled.

I was looking for that power containment slide too because I was wondering whether it dynamically clocked voltage too. From that slide it might do...that's a lot more validation than usual....

@Lightnix, as above: I expect that the default TDP would only be an issue in synthetic benchmarks where the clock may throttle. Something as a gamer, I wouldn't care about. However, someone does need to test if it does throttle on games hehe.
 
£285! That is a good price if indeed these are better than the 570s and only slightly behind the 580s. I will probably get one then!
 
That's why I wrote that the synthetic benches for the 6900's don't match reality of in-game performance. Perlin Noise was low precisely because the power limiter was in place. I bet if someone cranks it up 10% in power, scores shoot way up
 
so from the rumors now.... (neliz) the 6970 will be on par or 5-10% faster than 580 and this is with the power limiter at default!?? WOW. so just turn up the slider and POW!! another 10-15% on top of that???

Man this is crazy if true...
 
That's why I wrote that the synthetic benches for the 6900's don't match reality of in-game performance. Perlin Noise was low precisely because the power limiter was in place. I bet if someone cranks it up 10% in power, scores shoot way up

The benchmarks where the 6970 is being compared against 580; is the 580 at stock speed? If the powertune is increased further by 10% on 6970, wouldn't the 580 still beat it if it was overclocked to max aswell?
 
The benchmarks where the 6970 is being compared against 580; is the 580 at stock speed? If the powertune is increased further by 10% on 6970, wouldn't the 580 still beat it if it was overclocked to max aswell?

The difference is that it seems like the 6970 would be set to advertised clock speed when the slider maxed (basically downclocked before that) which would be comparable to the 580 at stock speed aswell.

By default the 6970 would be clocked at 780Mhz core (if comparable to the 6950 slide) with a 880Mhz rated chip. Then past 880 could be considered overclocking.
 
The difference is that it seems like the 6970 would be set to advertised clock speed when the slider maxed (basically downclocked before that) which would be comparable to the 580 at stock speed aswell.

By default the 6970 would be clocked at 780Mhz core (if comparable to the 6950 slide) with a 880Mhz rated chip. Then past 880 could be considered overclocking.

This is exactly what I was thinking aswell. So when the powertune slider is maxed it will bring 6970 to the stock speed of 580. Afterwards it all depends on which card overclocks better thus retaining the lead.
If this is indeed the case then an overclocked 6970 can be a serious contender to an overclocked 580.
 
Last edited:
This is exactly what I was thinking aswell. So when the powertune slider is maxed it will bring 6970 to the stock speed of 580. Afterwards it all depends on which card overclocks better thus retaining the lead.

Well the exact speed of the 580 is not relevant to the 6970, its designed to be as fast as they wanted but...
Yep, this is what the slide seems to indicate. We will see what happens when AMD releases CCC2 support for the 6900 series.

Perhaps the 6800 series also has unlockable potential as well :p
 
Last edited:
Oh boy now the heat seems to be turning on and 6970 isn't what we thought it to be with those initial benchmarks that we all have been seeing which reflects 'badly' on 6970. I think 6970 does have a lot of potential:).
 
Well the exact speed of the 580 is not relevant to the 6970

I can understand what you mean as they both have different architectures and for same speed, will give different results. In the end it all depends on which card gives the best performance when it reaches it's maximum potential :).
 
IF the information leaked is true then yes we haven't see a 6900 run at its natural potential yet. It's been so up and down there is no point in getting too eager about this as it may still be false and might not deliver a game changing size increase if true.

Lets keep our feet on the ground :)
 
IF the information leaked is true then yes we haven't see a 6900 run at its natural potential yet. It's been so up and down there is no point in getting too eager about this as it may still be false and might not deliver a game changing size increase if true.

Lets keep our feet on the ground :)

True. Although currently 580 appears to be the faster card but in 2 days time when the 6970 card is released we will know if it's still the case. Even if 6970 doesn't beat 580 at it's maximum overclock speed, I hope the performance gap is still a small one :).
 
Back
Top Bottom