GTX 1080 – What’s not being discussed

[Rroff]

Confirmed Nvidia were running fan at 100%

I had my doubts at the time hence my posts wondering what the actual sustained boost over time would be.

This is all rubbish.

8.2 Tflops is the base clock figure 8.9 T flops is the boost clock figure.

Also, it very much does put out 980 sli performance.

Taking the average sustained boost seen by reviewers the card is running at a theoretical peak of around 9.5-9.7TF in actual gaming - you really have to be desperate for a way to stick it to nVidia to complain about the TF figures heh - potentially on aftermarket cooling and overclocked you could be looking at way in excess of 10TF.

 

[Kaapstad]

http://www.overclock-and-game.com/news/pc-gaming/46-gtx-1080-what-s-not-being-discussed

I definitely think this rage is the work of an angry AMD Zealot.

Like any card the GTX 1080 does have a few weaknesses but unfortunately this badly written article has failed to find them.

 

[varkanoid]

1080 is the new king . all 980ti and TX owners will have to accept it.

its just a graphics card. Its not like it going to run for president.
who cares.

 

[Rroff]

its just a graphics card. Its not like it going to run for president.
who cares.

I wouldn't bet against it in a run for president with the current hopefuls.

 

[beany_bot]

This is all a load of cobblers.

Also, it very much does put out 980 sli performance.

Except... You are wrong.

 

[Rroff]

Really... the Arctic Panther uses 2x highly clocked Gigabyte GTX 980 G1... (Did I mention they are also watercooled?)

 

[beany_bot]

Really... the Arctic Panther uses 2x highly clocked Gigabyte GTX 980 G1... (Did I mention they are also watercooled?)

Sorry, does watercooling make a card faster...???
We all know that the 1080 doesn't clock for toffee and no WAY could it catch those 980's even clocked underwater. The gap is too big.

 

[Rroff]

Sorry, does watercooling make a card faster...???
We all know that the 1080 doesn't clock for toffee and no WAY could it catch those 980's even clocked underwater. The gap is too big.

The fact he is running them around 27% faster than the stock GTX980 boost clock might have some implications..........................

 

[beany_bot]

The fact he is running them around 27% faster than the stock GTX980 boost clock might have some implications..........................

Totally agree.
And the 1080 won't do that. Therefore. the speed you can get out of 2x 980s remains much more than a single 1080.
It's not rocket science....

 

[thebennyboy]

This thread is going the way of the dodo very quickly.

 

[Ayahuasca]

Totally agree.
And the 1080 won't do that. Therefore. the speed you can get out of 2x 980s remains much more than a single 1080.
It's not rocket science....

How do you know they won't?

Lets wait for the custom 1080's first that people can put under water, flash with custom bios and overvolt first. That would be a fair comparison.

 

[andybird123]

Totally agree.
And the 1080 won't do that. Therefore. the speed you can get out of 2x 980s remains much more than a single 1080.
It's not rocket science....

nice strawman, totally irrelevant to what NVidia actually claimed though

 

[Ayahuasca]

Also you've cherry picked results where SLI scales pretty much 100%.

 

[Rroff]

The test linked to in a couple of other threads with a hybrid cooler whacked on a 1080 showed an increase of 24.6% over the on paper boost clock versus the 27% the 980 is getting. Not really much in it. Which is kind of beside the point of the promotional stock versus stock claims.

 

[beany_bot]

The test linked to in a couple of other threads with a hybrid cooler whacked on a 1080 showed an increase of 24.6% over the on paper boost clock versus the 27% the 980 is getting. Not really much in it. Which is kind of beside the point of the promotional stock versus stock claims.

Stock for stock I don't know the results. Maybe the 1080 is better.
But I'm willing to put money on the fact that a max clocked 1080 wont beat max clocked SLI 980's. There isn't the room to play with that Maxwell had.

 

[Jono8]

Except... You are wrong.

against stock 980 sli obviously.

Not two watercooled g1's! What clocks are those 980's at?

 

[flopper]

http://www.overclock-and-game.com/news/pc-gaming/46-gtx-1080-what-s-not-being-discussed

The issue is Maxwell is dead for future dx12 titles mainly.
due to the way the tech was designed.
I wont be suprised that drivers are going to stop being updated without performance enhancement on maxwell also.
all to make the new card to look better...

You can wait "Some Nvidia fans are confident drivers will fix the issue, however we have been waiting around 9 months for those async drivers Nvidia stated they were working on." but it wont happen.

 

[Rroff]

Not two watercooled g1's! What clocks are those 980's at?

1545MHz I believe - not as high as I would have expected as I know a fair few people with 1500MHz clocks on like the WF3 air cooling.

 

[Ayahuasca]

But I'm willing to put money on the fact that a max clocked 1080 wont beat max clocked SLI 980's. There isn't the room to play with that Maxwell had.

I don't know how you can come to this conclusion yet.

Reference 980 and 980ti reviews had overclocks much more reserved than we finally saw with custom models.

If the 1080 is limited by power then one of the AIB's will find a solution.

 

[Jono8]

1545MHz I believe - not as high as I would have expected as I know a fair few people with 1500MHz clocks on like the WF3 air cooling.

Exactly. That is like comparing the 1080 to stock 980Ti sli performance, seen as a 980 at those clocks will be getting close to 980Ti performance.

Such a stupid comparison.