Possible GTX 880 benched in 3DMark Firestrike Extreme!

[3DVagabond]

Well, if GM204 is at ~400mm^2 then im expecting to be like the GK104 over GK110. That is, the GM204 will sacrifice GPGPU performance for Gaming and smaller die size when GM200 will be the high end SKU.

It's also supposed to be 256bit.

 

[Keysplayr]

It seems that GTX880 would be the successor to GTX680/770 and will be 28nm. Meanwhile GTX980 or "Titan ll" will be developed on 20nm node. The Kepler launch model seemed to work very well for them so why change it. That is of course if Nvidia can pull off once again their midrange taking on AMD's top end.

 

[DominionSeraph]

Wow, they're releasing a chip that won't even be faster than a 90 THz Voodoo 2? It's over. Nvidia is finished.

 

[raghu78]

It seems that GTX880 would be the successor to GTX680/770 and will be 28nm. Meanwhile GTX980 or "Titan ll" will be developed on 20nm node. The Kepler launch model seemed to work very well for them so why change it. That is of course if Nvidia can pull off once again their midrange taking on AMD's top end.

this logic of Nvidia's mid range taking on AMD's top end is rubbish. When AMD launched HD 7970 Nvidia replied with GTX 680. Both were high end and sold for same price and similar performance. Nvidia then brought their big die GK110 a year later. AMD replied with their Hawaii GPU.

 

[AtenRa]

Well Hawaii is AMD top end at the moment

 

[Keysplayr]

this logic of Nvidia's mid range taking on AMD's top end is rubbish. When AMD launched HD 7970 Nvidia replied with GTX 680. Both were high end and sold for same price and similar performance. Nvidia then brought their big die GK110 a year later. AMD replied with their Hawaii GPU.

Ok. Whatever makes you sleep better. I dont care either way. Just an observation of model numbers tells a different story. Especially how far back GK110 goes. Still Kepler. But you call it whatever you like.

 

[Genx87]

This could be real. But we don't have GPU clocks nor power figures. Would people be impressed if this thing was 30% faster than a stock 780 while using 50% less power consumption?

Given this will be on 28nm don't expect more than 20-30% more performance. Nvidia will be limited by the manufacturing process. When 20nm shows up then we will know the true potential of this design.

 

[Mand]

This could be real. But we don't have GPU clocks nor power figures. Would people be impressed if this thing was 30% faster than a stock 780 while using 50% less power consumption?

I don't expect a drop in TDP for the 880. I do expect them to translate the efficiency into higher performance.

 

[Genx87]

I don't expect a drop in TDP for the 880. I do expect them to translate the efficiency into higher performance.

I would be really surprised if they consume the same power as a 780 but only deliver 20-30% more performance given what the 750 was able to do.

 

[tviceman]

Serious question, not trolling: Do a lot of people just upgrade for the sake of upgrading/to have a newer/larger number in their PC?

A 25-28% faster 880Ti does nothing to fix the situation where we need more GPU power. Conversely, for 1080P, 780Ti slices through mostly everything else. That's why I just can't get excited even if the 880Ti is 28% faster for $550 over 780Ti.

For power users on 1440/1600P, 15-30% is still not good enough. For a real game changer beyond 290X/780Ti, we need 50-100% faster to make the gaming experience significantly better.

Is it unexciting to you, or are you just not interested? I'm not interested in GM204 (for the same reasons you state you don't want it), but it doesn't mean I'm unexcited for it. I'm not interested in the first retail release of Oculus, but it doesn't mean I'm not excited about it. New, faster GPU's at lower price points is ALWYAYS exciting to me.

Just because I'm planning to buy it doesn't mean I'm not excited. I feel like this pervasive "it doesn't benefit me, so it's crap" mentality is bred and perpetuated on forums, comments, and blogs.

How can anyone not be excited for a new high end at (hopefully) cheaper price points than what exist now? And if it's not for you, how can you not be excited for the potential of what today has and extrapolating that for what you may be getting tomorrow?

 

[guskline]

Interesting thread but RussianSensation's words really resonate with me.

I have an Oc'd 780 Classified underwater in my rig below (core 1306 and mem 1697) so performance is outstanding. I have 2 670 FTWs in SLI in my 3770k rig NS Sapphire R9 290 Tri-X in my 8350 rig.

My monitors are an Achieva 27" 2560 x1440 for the 780; Acer 27" 1920 x 1080 for the 670FTWs in SLI and a Hanns G 27.5" 1920 x 1200 for the R9 290. For the games I play all of my video cards give great fps.

The slow but steady progression of video cards is great for someone with an older card or a new build, but the upgrade urge just isn't there unless I would buy a 4k monitor.

 

[Phynaz]

Ok. Whatever makes you sleep better. I dont care either way. Just an observation of model numbers tells a different story. Especially how far back GK110 goes. Still Kepler. But you call it whatever you like.

So that's how we define top end now, by the model number?

 

[tviceman]

So that's how we define top end now, by the model number?

Top end is defined by performance. If it's the fastest on the market, the it's the TOP END.

What about this is so hard for people to understand?

 

[rtsurfer]

Top end is defined by performance. If it's the fastest on the market, the it's the TOP END.

What about this is so hard for people to understand?

So you mean to say the 680 was Top End.?

 

[Grooveriding]

I feel as if we should go back to the 7970 launch threads and see if we are hearing revisionist history here.

I tend to look at the entire lineup of cards released on the same architecture and node as the full gamut of the generation. That is how it has traditionally been until 28nm. Things are murky now with what were typically mid-range cards being released as high end and they're getting murkier now with what looks to be 4 years on the same node for GPUs. At the point where the new architecture is still coming out on the same node we got 3 years ago

 

[Mand]

So you mean to say the 680 was Top End.?

Yes.

A newer card, released later, was better, which meant that the 680 that used to be top end no longer was top end. This should not be shocking.

 

[chimaxi83]

Top end is defined by performance. If it's the fastest on the market, the it's the TOP END.

What about this is so hard for people to understand?

It's hard to understand for people with agendas, that's all. Calling an Nvidia card mid-range (even though it's their fastest available) and having it compete with AMD's top end is a "win". 7970 v 680 all over again.

 

[rtsurfer]

Yes.

A newer card, released later, was better, which meant that the 680 that used to be top end no longer was top end. This should not be shocking.

I have no issue with that.
But that is not what Keysplayr is arguing.

To him 680 was Mid Range & it surpassed AMD's Top End 7970 in Performance.

 

[Mand]

I have no issue with that.
But that is not what Keysplayr is arguing.

To him 680 was Mid Range & it surpassed AMD's Top End 7970 in Performance.

Perhaps, but I was referring to the people who say that you shouldn't get an 880 because the "top end" GM200 will come out later.

There will always be something better that will come out later.

 

[Atreidin]

There's top end for the market, which changes frequently, and there's top end for a specific generation of chips from a specific manufacturer. When people argue about this crap they repeatedly confuse the two, sometimes purposefully taking advantage of the ambiguity inherent in the conversation in order to suit an agenda.

 

[Atreidin]

I'd also add that most gamers most likely don't give a crap about the tech in their video card. I haven't seen any real studies on this, but I don't know any gamers who get outraged or even give any consideration to the generation of chips they are buying. Speed and price are the only two factors considered. All they care about is playing games, and the GPU tech is just a means to an end. They don't care if somebody considers what they have "mid range". If it's the fastest they can buy for the money they are willing to spend, they don't care what someone else calls it, because they are too busy playing games. If a faster, better card for the same money is coming in 1-2 months they might delay a purchase, but whether or not it is from the same generation isn't a factor.

 

[moonbogg]

Hmm. Fastish but derp.

 

[Rvenger]

Guys, let's get back on topic please.

-Rvenger

 

[2is]

I'll wait for game benches before getting excited or disappointed. I haven't benched using 3dmark in ages and don't even know what is or isn't a good score, which is probably for the best since I don't play 3dmark.

 

[tviceman]

So you mean to say the 680 was Top End.?

At the time of it's release yes. It was Nvidia's fastest card until Titan came out. Aren't the fastest cards considered to be top end? Just like Tahiti was, at the time of it's release.