Kitguru : Nvidia to release three GeForce GTX 800 graphics cards this October

Page 16 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

AtenRa

Lifer
Feb 2, 2009
14,003
3,361
136
Back to 3dmark11 performance and extreme. 3dmark11 performance preset is 720p from what I remember (someone refresh my memory), while extreme tests high resolution. Either the 880 is increasing its performance gap over the 780 as resolution increases, or the CPU testbed is affecting / inflating the guru3d score for the performance test.

Yes Performance is at 720p and Xtreme at 1080p. Using the X mode is heavily GPU bound.

Performance preset (P)
This preset is designed for benchmarking with a moderate load on the graphics card. The benchmark runs in
1280 x 720 (720p) making it suitable for most DirectX 11 capable gaming PCs, though at launch high end
hardware may be required to achieve a fluid frame rate.

Extreme preset (X)
This preset is designed for benchmarking with a very heavy load on the graphics card. The benchmark runs in
1920 x 1080 (1080p). The Extreme preset extends the lifetime of the benchmark by representing the likely
loads used by high end games in years to come. At launch it offers enthusiasts a suitable benchmark load for
competing at the extreme end of system performance.
 

Cloudfire777

Golden Member
Mar 24, 2013
1,787
95
91
Oh they are coming VERY soon
AIDA64 got a new update

  • BWCT LCD support
  • Crystalfontz CFA-53x, CFA-63x, CFA-73x LCD support
  • POS VFD support
  • Trefon 128x64 LCD support
  • improved LCD contrast configuration
  • improved bitmap display protocol on Matrix Orbital GLK, GLT LCDs
  • active page LED indicators on Matrix Orbital GLK, GLT LCDs
  • HW Monitoring / Logging / log started, log finished information
  • support for OpenGL 4.5
  • preliminary GPU information for nVIDIA GeForce GTX 880 (GM204)
    [*]preliminary GPU information for nVIDIA GeForce GTX 870 (GM204)
  • motherboard specific sensor info for Asus B85M-E/DASH, B85M-F, B85M-F R2.0, Q87M-A/TYCO/SI
  • fixed: System Stability Test / messaging issues
  • fixed: Sure Edition II and Edition III LCD support
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
Stop using AMD's oudated pre-GCN architectures. In Battlefield 4 a GTX580 is 35% faster than the 6970. On the other hand a 7970GHz - available 20 months later - is only 50% faster with 28nm and a much higher core clock.

Facepalm for the nth time. So now I have to ignore HD5870, HD6970 because you said so? Guess what many PC gamers had purchased HD5870 and HD6950/6970 and they upgraded to future NV/AMD cards. Don't like it, too bad. You can't choose to ignore previous generation of GPUs because you feel like it. Again, just because you don't buy AMD cards, doesn't matter to the rest of us who are brand agnostic. The fact is, you ignored perfectly valid comparisons such as GTX480 -> 780 or 580 -> 780Ti or GTX560Ti -> GTX680 which all brought 90-100% increases in performance in 3 years or less.

So again, you failed to provide any counter to my point other than aimlessly attacking both the 5870 and 6970 by cherry-picking some game benched at particularly settings that showed 480/580 outperforming the 5870/6970 cards by outlier % amounts. Fact is, 480 and 580 were only 17-18% faster on average and far less at 1600P. When discussing generations as a whole we don't just pick 1-2 games that shows outlier performance but look at the average itself since gamers play a wide variety of games. There is no need to cherry-pick 1 particular title and try to portray that as the average performance difference between 5870 and 480 or 6970 and 580 or it makes your post look very biased.

If you look at average performance, GTX580 doesn't beat HD6970 by some ludicrous 30-40% metric you keep claiming -- 17% at 1080P in the latest games. Therefore, it's perfectly valid to look at the performance jumps from 6970 to 7970Ghz or from 6970 to R9 290X/780:
http://www.computerbase.de/2013-05/nvidia-geforce-gtx-780-test/3/

Your post ignores reality because many gamers upgraded in such a way as 5870 and 6950 unlocked to 6970 were very popular cards.

For someone jumping from HD6970/580 to GTX780 Ghz/R9 290X (regardless of which card), the leap in performance is ~ 2x in demanding titles/modded games. Now you will claim that I just "cherry-picked" these 2 titles but I didn't since they portray what is roughly the performance increases on average in dozen of titles between generational jumps.




Again, you are trying to downplay the massive performance leaps Gamer X experienced by upgrading from flagship AMD or NV cards in the last 3 years by purposely cherry-picking some titles where AMD flagship cards perform the worst relative to NV's flagships while ignoring all the other titles where the reverse is true, as well as ignoring that even going with NV path only during the last 3 years ensured that this games got at least 2x the performance increase. The context is that if someone is upgrading from 680 OC/7970 OC to 880 OC, will get they 90-100% increase in performance? You keep saying yes but in the context of where 680 OC/7970 OC are vs. 780Ti, it would mean 880 would beat 780TI by > 30%.

If you still don't get the picture, move on. It seems no matter what the topic of discussion is you will always defend NV at all costs as if you have vested interest in the firm. For the rest of us, we will judge 870/880 according with previous increases in price/performance and absolute performance against existing cards.
 
Last edited:

Elfear

Diamond Member
May 30, 2004
7,116
695
126
So if the scores from Coolaler are correct and the 870 is right around the same performance as the 780, how likely is it that the 880 is much faster than the 780Ti? For example, if the 880 is 15% faster than the 780Ti, it would be ~42% faster than an 870 (since the Ti is ~25% faster than the vanilla 780).

When was the last time the *80 series card was 40%+ faster than the *70 series card?

Seems like too big of a gap to me.
 

alcoholbob

Diamond Member
May 24, 2005
6,271
323
126
So if the scores from Coolaler are correct and the 870 is right around the same performance as the 780, how likely is it that the 880 is much faster than the 780Ti? For example, if the 880 is 15% faster than the 780Ti, it would be ~42% faster than an 870 (since the Ti is ~25% faster than the vanilla 780).

When was the last time the *80 series card was 40%+ faster than the *70 series card?

Seems like too big of a gap to me.

If it has a high stock core speed its possible. It may be the overclocked gap between 870 and 880 is 20-25% but they just have a high base clock for the 880 relative to 870 in order to beat the 780ti in benchmarks. Although if that were the case it would be likely an overclocked 880 would lose to an overclocked 780ti.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
So if the scores from Coolaler are correct and the 870 is right around the same performance as the 780, how likely is it that the 880 is much faster than the 780Ti? For example, if the 880 is 15% faster than the 780Ti, it would be ~42% faster than an 870 (since the Ti is ~25% faster than the vanilla 780).

When was the last time the *80 series card was 40%+ faster than the *70 series card?

Seems like too big of a gap to me.

Very unlikely I say. I think we should be happy if the GTX880 matches the GTX780TI at reduced price and power consumption. All assuming that GTX870=GTX780.
 

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
So if the scores from Coolaler are correct and the 870 is right around the same performance as the 780, how likely is it that the 880 is much faster than the 780Ti? For example, if the 880 is 15% faster than the 780Ti, it would be ~42% faster than an 870 (since the Ti is ~25% faster than the vanilla 780).

When was the last time the *80 series card was 40%+ faster than the *70 series card?

Seems like too big of a gap to me.

If the scores from coolaler are correct, then the following scenarios will play out:

1) Perf / $ will improve significantly and Nvidia will tout this along with perf/watt. People with 780's / 290's or better will have no real incentive to upgrade. Nvidia will release GM200/210 a few months later as the GTX 880 TI and possibly GTX Titan M.
2) The GTX 870 is a heavily cut down GM204 chip, and GM204 will also occupy the GTX880 and GTX 880 TI slot, scaling up to 15-20% faster than GTX 780 TI.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,108
1,260
126
It's seems evident what we are going to see is price/perf and perf/w improvements, but not new performance levels. With these quite possibly genuine benchmark leaks and the specs on 880, I think it will actually be slower at 4K, surround and 2560x resolutions and about the same or maybe slightly faster at 1080p.

It will make an impressive 890 dual gpu card though. I think we're bound to see one of those release with the power consumption improvement.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
This too is an issue. It's the reason why I still think that single threaded performance is the most important thing. A lot of existing software is still single threaded.

The other question is, how much can be put into parallel. Some things probably cannot be. Amdahl's argument comes into argument here.


@Blackened and Noxious
Are we in agreement or do you disagree with this post?

I think both single and multi are important. Obviously single threaded performance plays out in even better performance in high thread count CPUs such as Xeon's, which would be to intel's benefit. But multi threaded performance is more complex than simply snapping your fingers and making it happen, I believe.

I think people are quick to blame devs for the issue, however, from the software developer blogs i've read in the past...the basic issue is that for CPUs, the more threads you have, it becomes excruciatingly difficult to program for. And therefore, the more threads you have, the less efficient things become with more threads. So you can have a 12 thread CPU and yes, you will see appreciable gains but typically what you'll see is that each that has far less than 100% utilization. That's the entire issue. Think of it as diminishing returns.

In the ideal world, everything would be programmed to use multi threaded CPUs, but in the real world,...it's not that simple from my understanding. If it were a simple snap your finger fix, man, CPUs would be progressing like crazy. But sadly it's more on the software end of things, and I wouldn't necessarily blame devs for this. There are real problems coding for multiple threads and the fact of the matter is, when considering the installed base of dual core ultrabooks and what not, 8-12 thread CPUs aren't all that common for their user base. Unless it's a strict scientific or productivity application.

The key difference is, like noxious said, GPUs specialize in parallel work. With this being the case, GPU cycles won't really slow down unless outside factors such as node transitions and a shift in the desktop market takes place. For CPUs? Not so much great at parallel workloads. You can have many cores and many threads...but it's a complex issue on the CPU side, basically, it's both a lack of software issue in addition to being a complex programming issue.

So I don't see GPU performance slowing down, not in the next 3-4 years at least. Beyond that? Who knows. Maybe node transitions will come to a standstill. I certainly hope that doesn't happen.
 
Last edited:

Elfear

Diamond Member
May 30, 2004
7,116
695
126
Very unlikely I say. I think we should be happy if the GTX880 matches the GTX780TI at reduced price and power consumption. All assuming that GTX870=GTX780.

Agreed.

If the scores from coolaler are correct, then the following scenarios will play out:

1) Perf / $ will improve significantly and Nvidia will tout this along with perf/watt. People with 780's / 290's or better will have no real incentive to upgrade. Nvidia will release GM200/210 a few months later as the GTX 880 TI and possibly GTX Titan M.
2) The GTX 870 is a heavily cut down GM204 chip, and GM204 will also occupy the GTX880 and GTX 880 TI slot, scaling up to 15-20% faster than GTX 780 TI.

Ya, doesn't make sense to have such a huge gap between their #1 and #2 cards. I see them going for better perf/$ this time around which will be good for consumers.

As an enthusiast I'd rather see some monster performance gains but with AMD currently silent on new cards, Nvidia doesn't have a whole lot of incentive to go for broke. That plus 20nm doesn't appear ready for prime time yet.
 

rgallant

Golden Member
Apr 14, 2007
1,361
11
81
Last edited:

Deders

Platinum Member
Oct 14, 2012
2,401
1
91



What am I missing? The 780 on Guru3D runs 12,100.

A though occurred to me that these might be 860 results.

When I said 780 performance level, it's in that range, you could push the same amount of pixels to near the same performance levels.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
With Samsung launching the Exynos 5430 on 20nm, all of these tech analysts just got proven wrong that no one besides Apple would have 20nm this year. Wow, and I trusted the professional research of those guys and their research papers. What do these people get paid for?

http://www.anandtech.com/show/8382/samsung-announces-exynos-5430-first-20nm-samsung-soc

This gives more hope that NV could either do a partial 20nm launch for 800 mobile or even be able to go 20nm for the desktop 870/880 parts.
 

Borealis7

Platinum Member
Oct 19, 2006
2,914
205
106
if 870 ~= 780 , i'm skipping that generation until 9XX series. not worth the upgrade in my opinion.
i might go for a used non-reference R9-290 if the price is right for the 20% upgrade.
 

CrazyElf

Member
May 28, 2013
88
21
81
With Samsung launching the Exynos 5430 on 20nm, all of these tech analysts just got proven wrong that no one besides Apple would have 20nm this year. Wow, and I trusted the professional research of those guys and their research papers. What do these people get paid for?

http://www.anandtech.com/show/8382/samsung-announces-exynos-5430-first-20nm-samsung-soc

This gives more hope that NV could either do a partial 20nm launch for 800 mobile or even be able to go 20nm for the desktop 870/880 parts.

The issue is that Samsung has its own fabs. I presume that their transition to 20nm went reasonably smoothly if they are launching now.

Nvidia makes their chips at TSMC - and GPUs for desktop are a part of the high power process, which is skipping 20nm, or at least last I heard. That may change later in 2015 though.


The key difference is, like noxious said, GPUs specialize in parallel work. With this being the case, GPU cycles won't really slow down unless outside factors such as node transitions and a shift in the desktop market takes place. For CPUs? Not so much great at parallel workloads. You can have many cores and many threads...but it's a complex issue on the CPU side, basically, it's both a lack of software issue in addition to being a complex programming issue.

So I don't see GPU performance slowing down, not in the next 3-4 years at least. Beyond that? Who knows. Maybe node transitions will come to a standstill. I certainly hope that doesn't happen.


I suspect that node transitions will come to a slowdown. The reason is because each new node seems to need exponentially more money to make and the benefits are diminishing per transition. I wish this were not the case, but the evidence right now suggests that it might be.
 

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
With Samsung launching the Exynos 5430 on 20nm, all of these tech analysts just got proven wrong that no one besides Apple would have 20nm this year. Wow, and I trusted the professional research of those guys and their research papers. What do these people get paid for?

http://www.anandtech.com/show/8382/samsung-announces-exynos-5430-first-20nm-samsung-soc

This gives more hope that NV could either do a partial 20nm launch for 800 mobile or even be able to go 20nm for the desktop 870/880 parts.

EH.... It's Samsung's fab, not TSMC's. Neither AMD nor Nvidia are creating chips at Samsung's foundries. And for all we know, Apple may still get their 20nm SoC on the market first. They are announcing their new lineup in 3 weeks, and it's never long after their announcement when products hit the shelves.

Don't get your hopes up. I think Nvidia is entirely skipping 20nm.
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
I think its certainly reasonable to expect a small 20 nm chip for mobile (GM107 successor).
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
EH.... It's Samsung's fab, not TSMC's. Neither AMD nor Nvidia are creating chips at Samsung's foundries. And for all we know, Apple may still get their 20nm SoC on the market first. They are announcing their new lineup in 3 weeks, and it's never long after their announcement when products hit the shelves.

Don't get your hopes up. I think Nvidia is entirely skipping 20nm.

I am not. I am waiting for GM200 or faster. Personally, I would rather NV wait 6 more months to launch 880 on 20nm with 30% more performance rather than on September on 28nm. Besides Witcher 3, there is no game coming out this year that I would upgrade for anyway.
 

CrazyElf

Member
May 28, 2013
88
21
81
Sighs ... that's the real botteneck at the moment. There are a few mods that I play that can fully saturate a 290, even at 2560x1440. But other than that, yep, the real bottleneck is the lack of GOOD (as in engaging, challenging, with a great deal of depth) games that really push the envelope.

Things might be different if I were on a 4K monitor. Off topic, but did AMD ever fix the issues with the r9 290s and 4k SST monitors?
 

Mand

Senior member
Jan 13, 2014
664
0
0
I am not. I am waiting for GM200 or faster. Personally, I would rather NV wait 6 more months to launch 880 on 20nm with 30% more performance rather than on September on 28nm. Besides Witcher 3, there is no game coming out this year that I would upgrade for anyway.

I still don't see how people can be so sure it won't be 20nm in September.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |