Overclock.net banner
1 - 17 of 17 Posts

·
Registered
Joined
·
569 Posts
Discussion Starter · #1 ·
Quote:
Introduction

A few days ago we introduced you to the concept of of frametime tests, a more accurate way of gauging graphics card performance in video games. Today we are applying the new test for the first time, and with the new data we will compare the AMD Radeon HD 7970 GHz Edition to the Nvidia GeForce GTX 680.

If you want the best of the best in the area of PC graphics capabilities, there aren't that many options out there and you will quickly narrow down your shortlist to include these two cards. The AMD Radeon HD 7970 GHz Edition is available for an average of £374 or € 400, and the Nvidia GeForce GTX 680 costs an average of £404 or € 460.

When we originally tested the 7970 GHz Edition, we concluded that the AMD card is faster in most benchmarks compared to the GeForce GTX 680. And when you take the lower price into consideration, it's clear that AMD gives you the best value for your money.

So, will this conclusion hold up in the light of new frametime data? We tested the two cards in five recent video games to find out.
http://uk.hardware.info/reviews/3991/amd-radeon-hd-7970-ghz-edition-vs-nvidia-geforce-gtx-680-frametimes-review
 

·
Registered
Joined
·
3,146 Posts
I really think this is great, more review sites including frame latencies tests on GPU reviews will only result in good things.

Hopefully it becomes more of a focus for driver/hardware teams for both camps.
 

·
Premium Member
Joined
·
9,327 Posts
So different results from multiple sites...
 

·
Registered
Joined
·
3,128 Posts
one of the biggest flaws of graphics drivers from Red and Green have been they are always biased towards max fps and not average or frame time. The latter two are way more important to the user's overall experience
 

·
Premium Member
Joined
·
6,630 Posts
I don't know why the GTX 680 would spike like that with medium settings in BF3. This article matches my testing in FC3, BF3 and Hitman though. Glad that Max Payne 3 and Assassin's Creed 3 perform so well on AMD, didn't know about those games.
 

·
Registered
Joined
·
2,927 Posts
too bad these reviews aren't testing this year's most intensive title...Crysis 3

At the end of last year I looked back the last 6 months to see if my graphics card purchase was worth it.

At the time of purchase factory overclocked 670 SLI for $650 was very hard to beat. AMD definitely has won the value/perf crown at the end of the year with its aggressive price cuts/huge performance boost from drivers. I felt a bit shafted.

Then I remembered why I went with Nvidia....terrible experience with legacy drivers from AMD for years...I wanted to make sure if I were to go multi-GPU I want some security that it would be a relatively smooth experience. At the time Nvidia owners claimed SLI experiences generally is smoother than CFX but no one had the data to back that up.

Well now with all the frame data coming out....I now feel the current price premium for Nvidia is worth it.

Bottom line now backed up by the frame time data...if you want brute force and better value...go AMD...if you want smooth experience go Nvidia.

What do you all say? I hope this doesn't spark any more uncesssary flame/fanboy wars.
tongue.gif
 

·
PC Evangelist
Joined
·
48,304 Posts
So now somehow GTX680 is faster then HD 7970? Maybe the 1080p resolution?
 

·
Premium Member
Joined
·
9,327 Posts
Quote:
Originally Posted by ZealotKi11er View Post

So now somehow GTX680 is faster then HD 7970? Maybe the 1080p resolution?
Probably a combination of 1080p + low AA,faulty card or user error.Especially at fb3 7970 is faster than a gtx680 from the majority of reviews
 

·
Registered
Joined
·
1,460 Posts
Quote:
Originally Posted by Z Overlord View Post

one of the biggest flaws of graphics drivers from Red and Green have been they are always biased towards max fps and not average or frame time. The latter two are way more important to the user's overall experience
Well, previously in almost all of the GPU reviews/benchmarks, only the FPS was taken in account. So it made sense for the drivers to be optimized for FPS.

Last year, according to one AMD driver developer, he said that they incorrectly believed that the microstuttering on single GPUs was not common and just limited to a few games, and now they have to play the game of catch up to fix the stuttering.
 

·
Registered
Joined
·
2,067 Posts
Quote:
Originally Posted by twitchyzero View Post

too bad these reviews aren't testing this year's most intensive title...Crysis 3

At the end of last year I looked back the last 6 months to see if my graphics card purchase was worth it.

At the time of purchase factory overclocked 670 SLI for $650 was very hard to beat. AMD definitely has won the value/perf crown at the end of the year with its aggressive price cuts/huge performance boost from drivers. I felt a bit shafted.

Then I remembered why I went with Nvidia....terrible experience with legacy drivers from AMD for years...I wanted to make sure if I were to go multi-GPU I want some security that it would be a relatively smooth experience. At the time Nvidia owners claimed SLI experiences generally is smoother than CFX but no one had the data to back that up.

Well now with all the frame data coming out....I now feel the current price premium for Nvidia is worth it.

Bottom line now backed up by the frame time data...if you want brute force and better value...go AMD...if you want smooth experience go Nvidia.

What do you all say? I hope this doesn't spark any more uncesssary flame/fanboy wars.
tongue.gif
Nothing you said initially is wrong, but the conclusion makes a good point of saying that both cards offered a smooth experience (besides FC3). To me this is more in line with what we see as consumers. People can use all the graphs and testing in the world but implying that it means the 7970 is a stuttering mess is foolish.

Now I'm not saying you said that but I don't think you're giving up a smooth experience if you go with a 7970 just because of a 2ms difference in frametime.
 

·
Registered
Joined
·
1,664 Posts
Quote:
Originally Posted by twitchyzero View Post

too bad these reviews aren't testing this year's most intensive title...Crysis 3

At the end of last year I looked back the last 6 months to see if my graphics card purchase was worth it.

At the time of purchase factory overclocked 670 SLI for $650 was very hard to beat. AMD definitely has won the value/perf crown at the end of the year with its aggressive price cuts/huge performance boost from drivers. I felt a bit shafted.

Then I remembered why I went with Nvidia....terrible experience with legacy drivers from AMD for years...I wanted to make sure if I were to go multi-GPU I want some security that it would be a relatively smooth experience. At the time Nvidia owners claimed SLI experiences generally is smoother than CFX but no one had the data to back that up.

Well now with all the frame data coming out....I now feel the current price premium for Nvidia is worth it.

Bottom line now backed up by the frame time data...if you want brute force and better value...go AMD...if you want smooth experience go Nvidia.

What do you all say? I hope this doesn't spark any more uncesssary flame/fanboy wars.
tongue.gif
Well said! I generally agree with you. However I have only ever assembled an SLI setup before (9600gt's and then 680's) so I can't comment on the ATI experience. I have been using ATI cards in the 5x and 6x series, but own an nvidia powered laptop. I have had literally almost no problems with the ATI drivers in almost 5 years now and recommend them to everyone I know for desktop cards. I believe both companies make good products!

Dunno what I'm getting at besides sharing my 2 cents!
 

·
Registered
Joined
·
1,959 Posts
Quote:
Originally Posted by dmreeves View Post

Well said! I generally agree with you. However I have only ever assembled an SLI setup before (9600gt's and then 680's) so I can't comment on the ATI experience. I have been using ATI cards in the 5x and 6x series, but own an nvidia powered laptop. I have had literally almost no problems with the ATI drivers in almost 5 years now and recommend them to everyone I know for desktop cards. I believe both companies make good products!

Dunno what I'm getting at besides sharing my 2 cents!
Don't worry, someone who hasn't used an AMD card in the last 5 years will come in shortly to tell you that your opinion is rubbish!
thumb.gif
 

·
Registered
Joined
·
2,927 Posts
Quote:
Originally Posted by CaptainChaos View Post

People can use all the graphs and testing in the world but implying that it means the 7970 is a stuttering mess is foolish.
I'm not saying 7970 stuttering is out of this world

but 7950/7970 in CFX...? From all the recent frame time reports you'll have to colour me skeptical even if I don't have 1st hand experience. Perhaps it'll take some time to perfect this new methodology or as someone else in the other thread defending AMD said it can be fixed/reduced by tweaking...but I would personally be uncertain to go with AMD CFX if I were in the market after reading all this.
 

·
Registered
Joined
·
569 Posts
Discussion Starter · #16 ·
Quote:
Originally Posted by CaptainChaos View Post

Nothing you said initially is wrong, but the conclusion makes a good point of saying that both cards offered a smooth experience (besides FC3). To me this is more in line with what we see as consumers. People can use all the graphs and testing in the world but implying that it means the 7970 is a stuttering mess is foolish.

Now I'm not saying you said that but I don't think you're giving up a smooth experience if you go with a 7970 just because of a 2ms difference in frametime.
I have seen it mentioned that around 20ms is unnoticeable.
Quote:
Originally Posted by SKYMTL;5173347
You are reading the chart completely wrong.

The variation you see on the Crossfire solution is ALL below 20ms, making it literally imperceptible to the end user.

Honestly, without perspective people can take these charts and have them completely misrepresent the information being provided. A card can do the bloody hokey pokey all over the place below 20ms and it wouldn't make one lick of difference since the actual REAL WORLD experience would still read as smooth.
Quote:
Originally Posted by SKYMTL;5173350
Again, you are making a blanket statement.

Do rapid fluctuations affect gaming? YES. However, the microstutter you see is likely NOT below 20ms but rather higher than that. Typically when deltas of 10ms or more quickly fluctuate at 25ms and above, you'll see microstutter. Anything below 20ms and you'd either completely miss it or be VERY picky and go looking for it.

Let me reiterate:

- Below 20ms: unnoticeable
- Rapid, cyclical fluctuations above 25ms: you'll notice it
- Sudden "spikes" (non-cyclical) above 40ms: you'll notice it
- Grey area between 21ms and 25ms: debatable. Some will notice it while others won't.
Quote:
Originally Posted by SKYMTL;5173361
It isn't me who is defining "noticeable". Read any article or scientific paper about perceptive latency and you'll likely come to the same conclusion.

When I make a statement, don't think for a second that I haven't researched it to an extensive degree.

Links to some of said papers:

http://www.stuartcheshire.org/papers/LatencyQuest.html

http://www.ncbi.nlm.nih.gov/pubmed/16639613 (link to PDF @ source)

http://www.perceptionweb.com/abstract.cgi?id=p240749

If you have an account @ Sciverse / Science Direct, there are several other articles in their archives which discuss the same thing. Unfortunately, I can't post links to them due to copyright terms but you'll quickly get the idea after Googling a bit.

However, I will distill it down into plain speak: In extensive lab tests with and without human subjects, a time of 48 frames per second (ie: 20.83ms) was deemed to be the threshold at which the majority of people anbd test instruments saw completely smooth display images or refreshes.

One test actually went so far as to rapidly flash a lightbulb at variable speeds in order to demonstrate when people would see it "flicker". That threshold for 95% of the participants was at 50 flashes per second (ie: exactly 20ms).

Not sure how much more information anyone could be looking for here....
wink.gif
http://www.xtremesystems.org/forums/showthread.php?285164-PCPer-Frame-Rating-Part-3-First-Results
 

·
Registered
Joined
·
2,067 Posts
Fair enough. There are fixes/tweaks, like setting a frame limit but out of the box SLI seems to offer a smoother experience atm. I cant speak personally about CFX either, but I do have a second 7950 coming in on thursday. The last time I had a dual card setup however was SLI 470's and microstutter was noticeable in certain games. Nvidia has definitely done some tweaking to fix that. Hopefully AMD can do something similar.

EDIT: Thanks for that post Final8ty, that should help put some perspective on things!
thumb.gif
 
1 - 17 of 17 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top