Kepler Nvidia GeForce GTX 780

[GoldenTiger]

Well, end of discussion from me. I made my point clear and we just have different opinions here. What matters to me, doesnt matter to you.


Overall im happy about 500$ price tag. That should make the card i want cheaper too perhaps. Thats 7870, what im hoping to get soon.

You're welcome to your opinion, I just feel that it's based on faulty logic, . Regardless, I of course wish you the most enjoyment out of whatever you choose to buy. Clearly as you said there's a difference of opinion that isn't going to be swayed, so ending it here works for me.

[DilTech]

Thats wrong again. Min fps above 35 is what makes my gaming experience acceptable. Everything above it is not important since human eye can not tell the difference while using IPS screens like i do. Average fps is completely useless stat for me.

Minimum FPS IS important (most important aspect to me there is)--but, you can't merely go by the number alone. There's been several games where there's split second skips when entering new areas, and that ends up the minimum number; without seeing the performance over time minimum is just another digit on the board. Especially since, yes, sometimes that dip is a different number from one vendor to another (but neither one actually effects the game).

[zalbard]

So a card that dips to 20 fps once but is above 40 otherwise is worse than a card that dips to 35 often? You need to do some sort of "average time spent below x fps" analysis to get the true minimum.

Wait for Techreport's review. It'll show 99th percentile frame times, that should paint a fairly accurate picture.

[Cold Fussion]

Im not even going to make counter argument to this . You are just so wrong and i think you know it.

Actually I'm not. Let me show you.

Here is a short run through BF3 with my system, the min fps is 10, average is 35 and max is 74.



Y axis is render time in ms and x axis is render time. The largest spike is around 96ms which is an instantaneous fps of about 10. Now there are very few times when the fps is this low and while it is annoying it doesn't really constitute a tangible degrading of an experience because these small small fps are such a small percentage of the overall experience.

Now let's look at a boxplot with a 1.5 IQR whisker length:



Now this is painting an interesting picture of what is going on. We can already see we have an enormous amount of outliers in the data (roughly 10% of all frames are outliers by this definition) but we can see that near the top where our min fps is, there is a very, very low concentration of frames there, so small that it's insignificant compared the clumpings from 35-50ms. If we look at the percentiles of this data, we find the 99th percentile is 59ms, that means only 1% of frames are ever over 59ms. If we go further we find the 99.5th percentile is 64ms and the 99.9th percentile is 86ms. Have you noticed something yet? Less than 0.1% of the frames are anywhere near the maximum rendering time. To get the maximum rendering time you need to jump to the 99.99th percentile, 0.01% of the data is at the minimum fps. The minimum fps is just insignificant to the overall dataset because it is at the outer most extreme of it. To look at what really degrades performance you have to dig deeper into the data. So what happens if we through away all the data we considered and outliers before? We get this:


(This plot started from zero so we don't exaggerate the spikes to much)

Now this is really worrisome, we've thrown away all our outliers and the rendering time is still all over the place. This is a textbook example of microshuttering, and if you simply looked at min, average and maximum fps you would have absolutely no idea that the gameplay was this abysmal. Even looking at percentiles we do not see the full extent of just the level degraded gameplay this setup yields. Now I've done some further analysis on the this data set just to give you an idea of what is going on. What I've done is counted the number of frame spikes and the number of areas of sustain low fps (excerpt from the code comments so you can see the methodology)

	%% Frame render time difference of 10 ms between frames
    % This is to find frame spikes, large difference in rendering time
    % between consecutive frames is an indicator of stuttering type
    % gameplay.

	%% Find how many frames constitute a lowering of fps.
    % A lowering of FPS is taken as a frame time difference of atleast 8
    % milliseconds, and for the rendering time to be atleast 33.5 ms. A
    % steady rate of 33.5 ms isn't considering a lowering of fps.

Using this we can find that for 20% of the frames, the very next frame has a difference in rendering time of 10ms, and 2.6% of time there is a sustained period of low fps. This is exactly what we saw from the graphs, and that was while the bulk of the time the fps wasn't low, it was all over the place which is what really degrades the experience and not just the minimum fps.

TLDR: Using the average,min,max doesn't give you an indication of how the game is actually playing, it simply gives you the upper and lower bounds and a smoothed result.

[GoldenTiger]

http://www.tweaktown.com/reviews/462...iew/index.html

EDIT: Note they are using 300.83 drivers, not the newer 300.99 ones, for the GTX 680, looks fantastic though regardless.

[kromosto]

yes min fps is very important to me too but not as a number with in a fps/time graph. also about the min fps with the gpu boost it shows the total gpu bottlenecks but when cpu or the other part which makes the bottleneck gpuboost increases the fps. if i am not wrong this is the case. of course there are cases like no one is bottlenecking but software doesnt push gpu to much also gpu boost can happen but i am talking about the extreme cases. now lets say while cpu is bottlenecking then overclokinn gpu should not increase fps too much. but in metro 1080p we can see a huge max fps diff from 7970 but min is lower then it. so this doesnt make sense unless it is about a premature driver.

[Russian]

6,739 words later... I'm done.

You guys can read my review soon. :P

[GoldenTiger]

6,739 words later... I'm done.

You guys can read my reivew soon. :P

Wow, that's a lot of words.... . Looking forward to it!

[Russian]

Wow, that's a lot of words.... . Looking forward to it!

Damnit you caught my typo my ninjaedit was unsuccessful!

[GoldenTiger]

damnit you caught my typo my ninjaedit was unsuccessful!

rofl .

[The Jesus]

6,739 words later... I'm done.

You guys can read my review soon. :P

TweakTown already published :P

[mAJORD]

Tweaktown had no NDA :p

[The Jesus]

Tweaktown had no NDA :p

They should've pushed weeks ago, would've made this thread so much easier to argue

Anyone notice the power consumption? 680 comes neck and neck with an overclocked 7970, nice efficiency (no GPGPU even) Nvidia, lmao.

[Russian]

They should've pushed weeks ago, would've made this thread so much easier to argue

Anyone notice the power consumption? 680 comes neck and neck with an overclocked 7970, nice efficiency (no GPGPU even) Nvidia, lmao.

Uh... power consumption yeah. You'll have to wait.

Tweaktown went with their review with no help from NVIDIA, so not everything will be right.

[The Jesus]

Uh... power consumption yeah. You'll have to wait.

Tweaktown went with their review with no help from NVIDIA, so not everything will be right.

Nvidia is going to help every person who buys a 680 personally set it up to achieve desired results? I'm just pulling your leg. But it really shouldn't be more complicated than, plug it in, measure with same base at same point in time.

[trinibwoy]

Well, end of discussion from me. I made my point clear and we just have different opinions here. What matters to me, doesnt matter to you.

trinibwoy, I choose my card and game settings based on one thing only. In no point of time, FPS can not fall below 35 when i play. Simple as that.

Overall im happy about 500$ price tag. That should make the card i want cheaper too perhaps. Thats 7870, what im hoping to get soon.

It's your money

[Kristoferr]

Everything as expected, execpt power consumption tests, which have huge difference from what we are seen so far.

[Mechanical Man]

Uh... power consumption yeah. You'll have to wait.

Tweaktown went with their review with no help from NVIDIA, so not everything will be right.

I would expect no help from AMD nor nVidia when doing fair review.

[zerazax]

TLDR: Using the average,min,max doesn't give you an indication of how the game is actually playing, it simply gives you the upper and lower bounds and a smoothed result.

The answer is that all of the factors matter, but you need a frame of reference (a time trend) for it to make sense

I would expect no help from AMD nor nVidia when doing fair review.

Hahaha! So true

[RealTelstar]

GeForce GTX 690 (Mai)
vermutlich 3072 (1D) Shader-Einheiten, 256 TMUs, 64 ROPs, 2x 256 Bit DDR Interface, Spieleverbrauch vermutlich ~330 Watt, Performance geschätzt 500-530%

GK110 (August)
vermutlich ~2500 (1D) Shader-Einheiten, 512 Bit DDR Interface, Spieleverbrauch vermutlich 250-300 Watt, Performance geschätzt ~495%

This monster ... GK110 will be 5% less than a dual of the same generation? OMG!

http://www.3dcenter.org/news/vermutl...90-aufgetaucht

GK110 comes out much after august IMO, depending on AMD release
The performance is what makes me think whether should i get a 590 or wait longer... I think i need a bigger main monitor first, but no 120hz 30 inches anywhere around :/

[trinibwoy]

I would expect no help from AMD nor nVidia when doing fair review.

If you're only concerned with benchmark graphs yes but Tweaktown wasn't complaining about help with benchmarking. Check the first page of the review.

[The Jesus]

GK110 comes out much after august IMO, depending on AMD release
The performance is what makes me think whether should i get a 590 or wait longer... I think i need a bigger main monitor first, but no 120hz 30 inches anywhere around :/

We're working on that monitor (27" but w/e): http://www.overclock.net/t/1232084/y...#post_16766759

However, we run into the issue of not being able to do 120Hz at 2560x1440(1600) due to some limitation (I just emailed AMD about it) either from pixel clock rate or RAMDAC or something.

[villa1n]

Everything as expected, execpt power consumption tests, which have huge difference from what we are seen so far.

http://www.tweaktown.com/reviews/462...w/index17.html These numbers look good?

[ajaidev]

No support for Tweaktown from Nvidia is really a bad thing to do. I mean when you have enough sample and can retail the product why not give a sample to tweaktown?? Nvidia is not only cutting off a thumb here but also inviting others to do the same action as tweaktown did. If it was Intel they would have gotten a replay as to why they cant access the product or any other polite comeback. Our PR department is very polite when dealing with reviewers some are special other not so much but each reviewer is taken care of by the elite.

[GoldenTiger]

No support for Tweaktown from Nvidia is really a bad thing to do. I mean when you have enough sample and can retail the product why not give a sample to tweaktown?? Nvidia is not only cutting off a thumb here but also inviting others to do the same action as tweaktown did. If it was Intel they would have gotten a replay as to why they cant access the product or any other polite comeback. Our PR department is very polite when dealing with reviewers some are special other not so much but each reviewer is taken care of by the elite.

I agree, assuming this side of the story is true.