• 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.

** Official 680 review thread **

RavenXXX2 said:
According techpowerup there is 2% in it compared to a 6990 so not much in it v 6970 xfire IMO.
Click to expand...

That review is showing only 30 to 40 percent between the 6970 and 6990 across all resolutions. Im sure 6970 crossfire scales more than 40 percent. They are using too low res and its taking down the average.

Id guess more like 20percent behind at 1920 and behond..
 
ALXAndy said:
No don't be silly. The 7970 is a faster card once you overclock it.

Remember.. Avoid green fever !
Click to expand...

Agreed. I think anyone with a 7970 would be mad to switch to a 680 unless they need Nvidia, or it happens to run particularly well in the games they play. Both very good GPUs, nice for the 7970 to have real competition


ALXAndy said:
But of course. See also - Apple.
Click to expand...

Let's not start that one :p
 
ALXAndy said:
Wait. So you are telling me that a card that overclocks itself to put out higher numbers against a card that does not is fair?
Click to expand...

Surely you see that you could simply turn that argument on its head:


The 680 is capable of running at its boost clock with complete stability, and dynamically chooses lower clockspeeds where appropriate, in order to reduce power consumption. Shouldn't we then be comparing and judging the 680 only at its boost clock?



The answer to both questions (the one you pose and the one above) is "no". Nvidia have released a card with a new power-saving / performance adjusting technology. This technology is in the retail card, and is enabled by default. Therefore it must be considered in the review. It is a design feature of the card - NOT a user-defined and card-specific overclock.

The only "true" comparison is an 'out-of-the-box' stock comparison. That is the ONLY level of performance which is guaranteed - anything beyond that is a bonus. Yes it's valid to compare factory-overclocked cards as this performance level is guaranteed, and when factory overclocked 680s start to appear that will continue to be true.

Beyond these levels the amount of overclocking is never certain (... all cards perform differently), so no valid comparison exists. "Clock-for-clock" comparisons only make sense if both cards are capable of the same maximum clockspeeds, or are released at the same clockspeed, which is rarely the case. The only truly valid "overclocked comparison" would come if a sufficiently large and wide sample size could be tested, with each card clocked to (say) the 95th percentile (i.e the core and memory clock which 95% of cards can achieve with stability). Even then you have issues of cooling (case and environment), which could affect these numbers.

Out-of-the-box numbers are the only truly valid comparison. Overclocked comparisons are interesting and informative, though never scientifically valid.
 
Duff-Man said:
The 680 is capable of running at its boost clock with complete stability, and dynamically chooses lower clockspeeds where appropriate, in order to reduce power consumption. Shouldn't we then be comparing and judging the 680 only at its boost clock?
Click to expand...

GPU boost causing some stutter:

'Lots of trouble, when we look at the time spent on long-latency frames. What happened to the GTX 680? Well, look up at the plots above, and you'll see that, very early in our test run, there was a frame that took nearly 180 ms to produce—nearly a fifth of a second. As we played the game, we experienced this wait as a brief but total interruption in gameplay. That stutter, plus a few other shorter ones, contributed to the 680's poor showing here. Turns out we ran into this problem with the GTX 680 in four of our five test runs, each time early in the run and each time lasting about 180 ms. Nvidia tells us the slowdown is the result of a problem with its GPU Boost mechanism that will be fixed in an upcoming driver update.'

http://techreport.com/articles.x/22653/9

There you go ALXAndy, Nvidia's answer to AMD's faulty zero core issues!
 
Last edited:
Duff-Man said:
Surely you see that you could simply turn that argument on its head:


The 680 is capable of running at its boost clock with complete stability, and dynamically chooses lower clockspeeds where appropriate, in order to reduce power consumption. Shouldn't we then be comparing and judging the 680 only at its boost clock?



The answer to both questions (the one you pose and the one above) is "no". Nvidia have released a card with a new power-saving / performance adjusting technology. This technology is in the retail card, and is enabled by default. Therefore it must be considered in the review. It is a design feature of the card - NOT a user-defined and card-specific overclock.

The only "true" comparison is an 'out-of-the-box' stock comparison. That is the ONLY level of performance which is guaranteed - anything beyond that is a bonus. Yes it's valid to compare factory-overclocked cards as this performance level is guaranteed, and when factory overclocked 680s start to appear that will continue to be true.

Beyond these levels the amount of overclocking is never certain (... all cards perform differently), so no valid comparison exists. "Clock-for-clock" comparisons only make sense if both cards are capable of the same maximum clockspeeds, or are released at the same clockspeed, which is rarely the case.

The only truly valid "overclocked comparison" would come if a sufficiently large and wide sample size could be tested, with each card clocked to (say) the 95th percentile (i.e the core and memory clock which 95% of cards can achieve with stability). Even then you have issues of cooling (case and environment), which could affect these numbers.

Out-of-the-box numbers are the only truly valid comparison. Overclocked comparisons are interesting and informative, though never scientifically valid.
Click to expand...

Well put! Sums it up nicely. Seems a good feature to remain within the TDP and dynamically change the clock
 
You must log in or register to reply here.
Back
Top Bottom