390X coming soon few weeks

Tahiti with the HD7970 GHZ edition Re-launched to be competitive with the GTX 680 and separate from GTX 670 segment.... and the same card rebranded as R9 280X as to cheap out cost they changed a lot of VRM making unable the voltage control and VRM monitorization.. also blocked voltage control implementing that shit called PowerPlay that only work good in the R9 285. all of those things together made the GHZ Edition have a poor OC headroom. averaging 1100mhz up to 1200mhz..

Got it, so 30% overclocks from original silicon = bad.
While GM200/GM204 are seen as amazing overclockers for their ~15-20% overclocks.
 
That is a bit short-sighted.
It is clocked roughly the same as Hawaii and has an AIO.
To think it has no OC headroom is simply ridiculous.


I think its pretty easy to see the fury cards using 300 watts or more though :/ might not have that much room to overclock because of that and heat dissipation for the none water cooled versions.

The cards that we have seen have been the water cooled with 2 8 pin.
 
speaking strictly with air cooler under typical conditions, a typical GTX 970 with a 1050mhz base clock can overclock easy between 1500mhz - 1600mhz without have to add any voltage.. GTX 980 with a bit higher 1126mhz base clock are in the lines of 1450mhz- 1550mhz...and those are where numbers are inline of +20%...

but what about of TitanX at 1400mhz-1500mhz with that crappy cooler and basically null power target limit?.. even the [H] sample was at above 1400mhz, anandtech's sample were above 1500mhz.. still 15-20% overclock?..

and With the GTX 980TI the situation its even better as the lower count of shaders, and participation of Aftermarket coolers make it even better overclocker.. so we are talking being negative of a 40% overclock in the highest end 980TI and TITAN X. which are able to deliver 20-30% extra performance and remember those are 1000mhz base clock.. I never said GM200-204 are amazing overclockers but certainly they can have a large overclock headroom.. and that's a thing that happens since kepler..
 
But what are their actual in-game clocks?

those must be the actual in-game clocks.. as nvidia clocks are always ingame/bench.. than what are reported in GPU-Z for example.. isn't rare to see those above 1500mhz certainly..
 
You guys are taking the stock base clocks and doing the calculations based on the overclocked max boost?
Seriously?

absolutely nope.. .you were the first to talk about Base clock and blablabla.. "30% over original silicon" remember?.. ;)

nvidia cards clocks are taken directly in-game. as they boost normally higher than advertised speed out of the box..
 
those must be the actual in-game clocks.. as nvidia clocks are always ingame/bench.. than what are reported in GPU-Z for example.. isn't rare to see those above 1500mhz certainly..

Well anandtechs Titan X sample was avg ~1125mhz boosted when in-game.
Overclocked it was ~1300 for ~16% increase in performance.

Same with GTX980Ti.
It was boosting a bit over ~1150mhz on average in games and overclocked was ~1350-1380mhz.

So I guess Nvidia's boosting is still very confusing to people where they can't see what is actually going on with the cards.

absolutely nope.. .you were the first to talk about Base clock and blablabla.. "30% over original silicon" remember?.. ;)

nvidia cards clocks are taken directly in-game. as they boost normally higher than advertised speed out of the box..
Wow, confused much?

7970 was released with 925mhz, no boost.
Easily overclocking to 1100-1200mhz. Later silicon was doing over 1200mhz.
 
Well anandtechs Titan X sample was avg ~1125mhz boosted when in-game.
Overclocked it was ~1300 for ~16% increase in performance.

Same with GTX980Ti.
It was boosting a bit over ~1150mhz on average in games and overclocked was ~1350-1380mhz.

So I guess Nvidia's boosting is still very confusing to people where they can't see what is actually going on with the cards.

lol hahaha it seems that you really never had a nvidia GPU since kepler... even AMD said in their review

This pushes the GTX 980 Ti's clockspeeds up to 1326MHz for the standard boost clock, and 1477MHz for the maximum boost clock.

and even in their page they state a max boost clock of 1477mhz obviously limited seriously by the reference cooler and by the low power limit of reference card.. even they state in their OC page a out of the box boost clock of 1202mhz.. that still have to be taken into consideration.. or because the card automatically overclock can't be taken seriously?.. the better card, the better boost they can get out of the box and that still count as overclock..
 
lol hahaha it seems that you really never had a nvidia GPU since kepler... even AMD said in their review
You are wrong, I've owned and played with multiple Maxwell cards. I am well versed with GPUs from both AMD and Nvidia.
"even AMD said in their review" what?


and even in their page they state a max boost clock of 1477mhz obviously limited seriously by the reference cooler and by the low power limit of reference card.. even they state in their OC page a out of the box boost clock of 1202mhz.. that still have to be taken into consideration.. or because the card automatically overclock can't be taken seriously?.. the better card, the better boost they can get out of the box and that still count as overclock..
Yes, max boost. Not sustained boost.
Ever heard of suicide runs?
Do you consider those 24/7 overclocks?

You are arguing that Nvidia cards are seeing 40-50% more performance when overclocked... which is incorrect.
 
lol hahaha it seems that you really never had a nvidia GPU since kepler... even AMD said in their review



and even in their page they state a max boost clock of 1477mhz obviously limited seriously by the reference cooler and by the low power limit of reference card.. even they state in their OC page a out of the box boost clock of 1202mhz.. that still have to be taken into consideration.. or because the card automatically overclock can't be taken seriously?.. the better card, the better boost they can get out of the box and that still count as overclock..

Umm, going by that logic Hawaii is an amazing overclocker then. Because base clock is likely 727 MHz, and even in quiet mode the lowest boost is 907MHz, or a 25% overclock.
 
You are arguing that Nvidia cards are seeing 40-50% more performance when overclocked... which is incorrect.

i've said they receive 20-30% more performance when overclocked.. just that..

Umm, going by that logic Hawaii is an amazing overclocker then. Because base clock is likely 727 MHz, and even in quiet mode the lowest boost is 907MHz, or a 25% overclock.

Hawaii doesn't even have a base clock on reference models.. in aftermarket its a fixed 1000mhz in most models.. and AMD advertised up to 1000... so the clock its really by your logic 1000mhz or less.. nvidia state a base clock with a boost clock..
 
Hawaii doesn't even have a base clock on reference models.. in aftermarket its a fixed 1000mhz in most models.. and AMD advertised up to 1000... so the clock its really by your logic 1000mhz or less.. nvidia state a base clock with a boost clock..

Well you said it yourself, AMD advertised up to 1000, meaning the base clock is clearly below 1000.

I'm simply pointing out that going by your logic of "boosting over base = overclock", then Hawaii also "overclocks" well.
 
I was under the impression that those Nvidia overclocks didn't translate well when you played games like The Witcher 3 and would cause crashing issues or the card to thermal. Not that they aren't fine cards. Just that the 1500 was wishful thinking when you actually played games.
 
I was under the impression that those Nvidia overclocks didn't translate well when you played games like The Witcher 3 and would cause crashing issues or the card to thermal. Not that they aren't fine cards. Just that the 1500 was wishful thinking when you actually played games.

Not just an impression, I can bench up to 1580 on my 970 without any issue, Try playing witcher or shadows of mordor equals eventual crashing anything above 1500. Not always right away but the crashes do come. Multiple reports around the web of the issue, and that isn't even including all the driver crashes with the newer drivers and watching videos online. Nvidia is not the perfect company many seem to think they are. ;)
 
I find it funny that all these videocardz.com leaked benchmarks we see lately always include an OCed 980ti and OCed 980, and no other OCed cards, in order to show more NVidia cards at the top.

I suspect bias in their 'reviews'.
 
They can't OC a card they don't have.
Both VCZ and WCCF heavily favor AMD. Just read their comments.

Their articles will always be extremist in one direction or the other -- that's how you get clicks.
"Fury is the fastest GPU in the world!"
"Fury is the slowest GPU in the world!"
Clicky click click.
 
Haven't heard from any staff for a while... Bet they are busy with the 300 series and hopefully Fury.
Seem really quiet since Computex across the interweb... Maybe just vacation time.
 
Try reading some of the thread...lulz :rolleyes:

Well, to be honest you have to wade through a whole lot of shit to get any little nuggets of news in a thread like this. This is mostly a thread about people telling other people why they are wrong. Over 5000 posts long and not much solid information.
 
Well, to be honest you have to wade through a whole lot of shit to get any little nuggets of news in a thread like this. This is mostly a thread about people telling other people why they are wrong. Over 5000 posts long and not much solid information.
If it's too tedious to follow the thread for you, maybe wait until the card gets out to know about it?
 
If it's too tedious to follow the thread for you, maybe wait until the card gets out to know about it?

He's not wrong, and one more post repeating something that was already said isn't gonna ruin the thread any more than your shitty post. So be nice?
 
the music almost sounds like Purple Motion (Jonne Valtonen) who was one of the Future Crew back in the early 90s.

https://www.youtube.com/watch?v=cJF0vUtVpCM

If I'm not mistaken, there is a reason for that! :p

I lived in Sweden during the height of the demoscene days. I never made it to Future Crews Assembly in Finland, but I did go to The Party in Denmark once and a few lesser known ones in Sweden.

I remember Purple Motion and Skaven and all those guys (well, their work, I never actually met them). Even tried my hand at composing some stuff myself in their Screamtracker.

But that was a long time ago, back when I was sporting a 486sx 25 with 4MB of RAM overclocked to 50mhz.

You don't get too many 100% overclocks anymore :p

Most of my friends were on Amiga's back then, but I was always a PC guy. At least since 1991 when I was 11, and switched from the NES to PC :p
 
Last edited:
Well, to be honest you have to wade through a whole lot of shit to get any little nuggets of news in a thread like this. This is mostly a thread about people telling other people why they are wrong. Over 5000 posts long and not much solid information.

That's why you should just read my post history.
 
If you want to know what we know, then go to WCCF and read all of their Fury articles from the last 3 or 4 days... Done.
 
He wasnt in the credits, i looked.

Interesting. Maybe it was Skaven then?

Was Peter Hajba in the credits? (I can't check right now)

The sound is unmistakingly Future Crew demo party module entry...
 
Last edited:
cJLKhSk.gif

now there's a better reason to get a windowed case... to show off the Radeon Fury!

that guy's using Silverstone Aeroslots in the case!
 
Back
Top