Possible GTX 880 benched in 3DMark Firestrike Extreme!

Page 10 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
I've said it before and i will say it again, unless GM204 is at 20nm there is no way it can be faster at high resolutions with only 256bit memory, 32ROPs and less Texture units(160 ??) than GTX780Ti 384bit, 48 ROPs and 240 Texture units. Especially at 4K.

Can I use this as a quote in my signature?
 
Last edited:

AtenRa

Lifer
Feb 2, 2009
14,001
3,357
136
I don't think you're seeing this clearly. The GTX750Ti has a deficit at every aspect and still manages to stay withing spitting distance of GTX650TiBoost. with or without AA. OF COURSE a bandwidth deficit is going to impact performance when higher levels of AA are used especially at higher resolutions. Point is, Nvidia did a whole lot more with a whole lot less with 750Ti. You don't agree?

Have you really paid attention to the graphs ??? GTX650Ti Boost is way faster at 1080p 4x AA.

Also, you speak of GTX750ti has a deficit in every aspect, well that is what GM204 vs GK110 will also be.

less stream processors 2560 vs 2880 ??
256bit vs 384bit memory channels ??
32ROPs vs 48ROPs ??
160 Tex Units vs 240 Tex Units ??

Again, IF GM204 is GTX880 and IF it is made at 28nm i dont expect it to be faster than GTX780ti in high resolutions with AA filters.
 

SirPauly

Diamond Member
Apr 28, 2009
5,187
1
0
For clarification:

The reference GTX 650 ti boost clocks are 980 base and 1033 boost. The MSI GTX 650ti boost was 1033 base and 1085 boost in that review!
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
I was going to say that MSI 650Ti Boost is a slightly overclocked version against a reference clocked and overclocked 750Ti. That could very well bring the 750Ti closer than the 12% you propose. And yes, that is spitting distance regardless if spitting distance to you would be far less in this case.
 

CakeMonster

Golden Member
Nov 22, 2012
1,391
498
136

Their predictions are all over the map. Seems they have no clue but still want to churn out some text so they have to go with something that could have been generated with a scrip and contradicts itself.

GeForce GTX 880 is using very similar board design to GTX 770. GeForce GTX 880 is of course 780 successor, but it has more similarities to GTX 770 than GTX 780.

The difference between GTX 880 and GTX 770 boards is cosmetic. Thus, GM204 is basically a GK104 replacement. However, just because it’s a mid-range GPU replacement, It does not mean it will be slower than some GK110 cards. We do not know how fast GTX 880 is going to be, but it’s safe to assume it will sit above 780 and maybe even 780 Ti.

Gee, really? :rolleyes:
 

voodoo7817

Member
Oct 22, 2006
193
0
76
If they release the 860 at the same time as the 880/870, I'll probably get the 860. But if they only release the 880/870 at first, and they are reasonably priced for their performance level (let's say $500/$350 with 780ti/780 performance), I might jump up for one of them as my 1440p monitor will be able to use the extra powah. My guess is that NVIDIA knows that there are a fair amount of people like me so we won't see a $200-300 dollar 800-series card in 2014 unless AMD forces their hand.

I'm curious, what's the biggest performance spread based on the same chip in the last bunch of years of GPUs? For example, the GK104 is in the 760 and 770, right? So in this example, what's the lowest performing and priced GPU with a GK104 chip and what is the highest? I know different generations come into play (isn't GK104 also in 680?) and mess with price points a bit, but I wonder if there's anything we can learn about the 860/870/880 performance spread from looking at what NVIDIA has done in the past.
 

ocre

Golden Member
Dec 26, 2008
1,594
7
81
IMHO, the mid range GTX880 will be marketed as the flagship until Big Maxwell comes out possibly on 20nm. Big Maxwell (GM210 or whichever number it's assigned) will show up on 20nm later on and that will be the GTX980/Ti and Titan 2/ Black GPUs.

Exactly like we have now with the entire Kepler line. Only difference is all of that was on one (28nm) node. A mid range chip on 28nm at 430mm2 isn't that out of the question when we are literally at the tail end of a very very mature 28nm node process. A GTX 980 Die might be the same size on 20nm as a GTX 880 is on 28nm. I'm speculating of course but does my speculation sound at all unreasonable?

the GM204 die is quite large. Calling it a midranged die is inappropriate. It is physically impossible to have the gm200 on 28nm.

We are at the mercy of the node. And the gm204 will be one of the largest 28nm chips there was.

Also since Nvidia has known for some time that they wasnt going to 20nm this year, it stands to reason that the gm204 was intended to be the 880 for some time. They should have had enough time to tweak the architecture to fulfill this role. So while the gk104 might have been pushed up to the 680 by pushing up clocks and ram speed, the gm204 might have had some major architectural changes to take it up a notch. I would be really surprised if nvidia hadnt.

The difference in the gk104 and gm204 is that Nvidia has known for a long long time where the gm204 was gonna be positioned. The gk104 was more of rushed effort.