Review Index:
Feedback

Frame Rating Dissected: Full Details on Capture-based Graphics Performance Testing

Vsync and its Effect on Frame Rating – Does it fix CrossFire?

After publishing the Frame Rating Part 3 story, I started to see quite a bit of feedback from readers and other enthusiasts with many requests for information about Vsync and how it might affect the results we are seeing here.  Vertical Sync is the fix for screen tearing, a common artifact seen in gaming (and other mediums) when the frame rendering rate doesn’t match the display’s refresh rate.  Enabling Vsync will force the rendering engine to only display and switch frames in the buffer to match the vertical refresh rate of the monitor or a divisor of it.  So a 60 Hz monitor could only display frames at 16ms (60 FPS), 33ms (30 FPS), 50ms (20 FPS), and so on with a 120 Hz monitor could also being capable of 8ms (120 FPS), etc. 

Many early readers hypothesized that simply enabling Vsync would fix the stutter and runt issues that Frame Rating was bringing to light.  To test this we looked for a game that ran right around the 60 FPS mark in our in normal testing with Vsync disabled and then set about to re-run results with it on.  We are using a standard 60 Hz monitor with the goal of being able to test some 120 Hz capability soon after we figure out a final bug or two with our capture configuration. 

First up, let’s take a look at the NVIDIA GeForce GTX 680 and GTX 680 SLI and see what shows up.

View Full Size

Because the average frame rate per second graph averages out the frame times for a total of one second of time, the averages won’t quite be the straight lines you might have expected.  Looking at the GTX 680 SLI Vsync enabled results the only key item is that the frame rate doesn’t go above 60 FPS like it does with Vsync disabled.

View Full Size

The single card and SLI configurations without Vsync disabled look just like they did on previous pages but the graph for GTX 680 SLI with Vsync on is very different.  Frame times are only switching back and forth between 16 ms and 33 ms, 60 and 30 instantaneous FPS due to the restrictions of Vsync.  What might not be obvious at first is that the constant shifting back and forth between these two rates (two refresh cycles with one frame, one refresh cycle with one frame) can actually cause more stuttering and animation inconsistencies than would otherwise appear.

View Full Size

Based on our graph here we found that with Vsync enabled we had about 87% of our frames running at 60 FPS (16 ms) and 13% at 30 FPS (33 ms).  You might be curious how there could be 60 FPS frame rate so often with Vsync on but very few frames at 60 FPS with Vsync off, and the answer lies in the rate limiting caused by Vsync.  Because of the back pressure on the game engine caused by the longer frame times (30 FPS, 33 ms) from Vsync there is more time for the GPUs to “catch up” and render another frame at 16 ms. 

View Full Size

Our ISU graph on stutter potential tells the story in a more damning light; starting at the 30th percentile the Vsync enabled setup of GTX 680s in SLI are already running at much higher frame variances and it only gets worse as we hit the 60s, 80s and 90s.  At the 90th percentile we are seeing frame variances over 12 ms, which is nearly a complete monitor refresh cycle!

 

Now let’s see how the AMD Radeon HD 7970 results change.

View Full Size

Something interesting is already happening here – the Vsync enabled results from the HD 7970 CrossFire configuration are running at HIGHER average frame rates per second than with Vsync disabled!  The orange line clearly never hits the 60 FPS mark while the black line (Vsync) does. 

View Full Size

Without Vsync we clearly see the runts affecting the plot of frame times here on the HD 7970s in CrossFire but enabling Vsync does appear to eliminate them! 

View Full Size

With our observed frame rate we have the same results for the HD 7970 CrossFire as we did with our FRAPS results, indicating no dropped frames or runt frames.  Standard CrossFire mode still shows the horrible results we have come to expect from our analysis today.

View Full Size

Our Min FPS percentile graph shows us that we are running at 60 FPS (16 ms) 85% of the time and 30 FPS (33 ms) the rest.  Because our data here is based the observed frame rates and not the FRAPS frame rates, there is no correlation between the two CrossFire runs.

View Full Size

The ISU graph of stutter potential again indicates that the Vsync enabled option is introducing higher frame variances than we would like and it is doing it more dramatically and earlier than the GTX 680s in SLI. 

It does appear that enabling Vsync will help alleviate the runts issue seen with AMD Radeon cards in CrossFire but at the cost of much more frame variance and stuttered animation on games that previously didn’t exhibit that problem. 

Let's take a look at another example using CrossFire that has another particular set of circumstances.  I theorized that in a gaming scenario that bordered just under 60 FPS with a single GPU, we would still see problematic results when jumping to HD 7970s in CrossFire.  Take our Battlefield 3 2560x1440 testing: with only one HD 7970 we are running just under 60 FPS most of the time which would, with Vsync enabled, force the game to run at 30 FPS with 33ms frame times.  Ideally we would like to see that move from 33ms frame times to 16ms frame times when adding in another HD 7970 in CrossFire due to the extra performance pushing the card over 60 FPS steady.

View Full Size

Our FRAPS graphs looks how we would hope and expect real-world performance to look.  While the single HD 7970 ran at a non-standard frame rate when performance was under 60 FPS, towards the end (50 sec point) where it could, we see a flat line that is partially hidden behind the pink line.  That pink line represents CrossFire HD 7970s and by doubling the number of GPUs we expected to maximize performance at 60 Hz with Vsync enabled, and we have. 

View Full Size

Observed frame rates calculated by removing runts are showing the Vsync DISABLED results on the HD 7970s in CrossFire mirror what we have seen before with much lower performance.  However, the Vsync ENABLED results did not change! 

View Full Size

The somewhat complicated plot diagram of frame times indicates that at no time did the frame rate of the HD 7970 cards in CrossFire go below 60 FPS or above the 16ms mark - even though there are thousands of frames under 16ms (runts) when Vsync is disabled.  Not only that but performance over the single HD 7970 with Vsync enabled is improved - rather than having jumps between the 16ms and 33ms frame times, we are locked in at 16ms - matching the 60 Hz refresh of our panel. 

View Full Size

The minimum FPS percentile graphic shows the same story - the pink link representing the HD 7970s with Vsync turned on looks solid.

View Full Size

Notice as well that with a static 16ms frame time we see no frame time variance at all in our ISU graph indicating that the kinds of stutter we are searching for are not showing up at all.

How is this happening?  How is enabling Vsync 'fixing' the runts and frame time issues of CrossFire?  The secret lies in the inherent back pressure of vertical sync to pace the graphics card and AMD's CrossFire engines even against its own will.  By forcing the GPUs to only render one frame every 16ms (at the maximum), Vsync is able to force the GPU to pace itself in a way that it would otherwise not.  This doesn't happen in every game though as we saw in the Crysis results first, and there is a lot more testing that needs to be done with Vsync to make a firm decision.

 

NVIDIA has a couple of different solutions in the NVIDIA Control Panel that might help: Adaptive Vsync and Smooth Vsync.  Adaptive Vsync was released with the first Kepler GPUs last year and we found it to be very effective at reducing stutter while also eliminating tearing.  Smooth Vsync is a little known feature that only exists in the driver when SLI is enabled as it takes advantage of many of the same frame metering features that SLI uses.  It attempts to keep frame rates “settled” at a level until it decides it has enough horsepower to move up to the next frame rate option for an extended period of time.  It is a very dubious description at best and NVIDIA didn’t go into much detail on how they decide if they have enough GPU overhead remaining or how long that “period of time” really is.

View Full Size

I decided to run through the same Crysis 3 sequences at 1920x1080 on the GTX 680s in SLI with all four NVIDIA options enabled: Vsync off, Vsync on, Adaptive Vsync and Smooth Vsync. 

View Full Size

Our FRAPS based results show the same similar looking results for standard Vsync on and off, but the adaptive and smooth Vsync options appear to be fixed at 30 FPS with the occasional hiccup on the Smooth Vsync.

View Full Size

The plot of frame times is kind of confusing but the important data is to compare standard Vsync On to Adaptive and Smooth.  With the exception of the 6 or so spikes on the smooth configuration the frames are basically fixed at 33 ms, resulting in a perfectly smooth gameplay experience but at the expensive of limiting performance. 

View Full Size

The observed FPS doesn’t change at all.

View Full Size

Another view here shows the same thing with a fixed frame rate of 30 FPS for adaptive and smooth Vsync options.

View Full Size

NVIDIA’s Adaptive Vsync shows basically 0 variance and only very minimal variance on the Smooth Vsync option at the 96th percentile.  So even though performance is lower on average, the experience is smoother.

 

NVIDIA’s additional Vsync options are definitely a strong point in favor of its technology though the Smooth Vsync only exists on SLI configurations.  I have been told that they were considering adding it to single graphics card configurations and I certainly hope they do as it adds some significant value in the same way Adaptive Vsync and Frame Rate Limiting do.

For both NVIDIA and AMD multi-GPU solutions with standard Vsync, enabling it definitely changes the story.  NVIDIA’s cards pretty much perform as we expected but for CrossFire we didn’t really know what expect with the various visual concerns.  It does appear that the runts problem was at least mostly solved with the enabling of Vsync though to be clear we are only testing a couple of game at this point – much more needs to be done. 

However, enabling Vsync creates a whole host of other potential issues that gamers have to deal with.  Even though the goal of removing visual tearing is met with the option turned on, you do add latency to the gameplay experience, as much as 60ms in some cases, from input to display.  Putting back pressure on the GPU pipeline, for both NVIDIA and AMD, means that some frames are going to be running behind schedule or behind the input timing of the game itself.  Many gamers won't want to deal with those kind of input problems and that is why many still play games with Vsync disabled.  Turning on Vsync does help AMD's CrossFire performance but it isn't the final answer just yet.

March 30, 2013 | 10:07 AM - Posted by Luciano (not verified)

About various vsync methods:
They're not the same code nor are available through the same ways.
But they are the same methods and persue the same results.
SLI and Crossfire are not the same thing...
But...

March 30, 2013 | 10:25 AM - Posted by John Doe (not verified)

This site and the people of it suck and choke on AMD's dick every single day.

They've recently gotten a Titan and have been shilling it so far but that's all sadly, compared to how much AMD shilling is going on on here.

Suckers.

March 30, 2013 | 11:48 AM - Posted by bystander (not verified)

You should read this:
http://www.tomshardware.com/reviews/graphics-card-benchmarking-frame-rat...

"You take the red pill - you stay in Wonderland, and I show you how deep the rabbit hole goes."
- Morpheus, The Matrix

"As a rule, human beings don't respond well when their beliefs are challenged. But how would you feel if I told you that the frames-per-second method for conveying performance, as it's often presented, is fundamentally flawed? It's tough to accept, right? And, to be honest, that was my first reaction the first time I heard that Scott Wasson at The Tech Report was checking into frame times using Fraps. His initial look and continued persistence was largely responsible for drawing attention to performance "inside the second," which is often discussed in terms of uneven or stuttery playback, even in the face of high average frame rates."

March 30, 2013 | 12:15 PM - Posted by John Doe (not verified)

My statement wasn't at Scott or whoever the fuck you are talking about that I don't give a rats ass about.

It was at the runners of this site.

THIS site, you're currently commenting on, fucking SCREAMS AMD ALL OVER. Just take off your AMD shirt already. Sigh.

March 30, 2013 | 01:33 PM - Posted by billeman

Please tell me there's a way do discard the (not verified) comments :)

April 3, 2013 | 11:47 PM - Posted by Jeremy Hellstrom

I could, but ...

March 30, 2013 | 06:38 PM - Posted by tidsoptimist (not verified)

This is a fascinating and quite informative bunch of articles.
Still i'm having some doubt, but as "bystander" stated above it's hard for a person to get challenged so hard in their beliefs.
This will make it hard for other sites to do as deep reviews as you do though and I hope you somehow start using open source on the code used so everyone who thinks you are payed by either "team" can check it out and even do some of the tests themselves with some modifications depending on the hardware used.

One thing I would like to know is if three cards would make any difference at all? I know its even more rare for people to have three cards and if you would use your conclusion then it wouldn't change much.
And this splitting you use, it is a passive one or is it somehow doing something to the stream?

Thank you for the informative articles.

March 31, 2013 | 02:34 PM - Posted by Luciano (not verified)

It would:

http://www.tomshardware.com/reviews/radeon-geforce-stutter-crossfire,299...

April 9, 2013 | 05:45 PM - Posted by tidsoptimist (not verified)

Thanks for the link!

trifire has had its benefits in the past as well thats why i started thinking about it..

March 30, 2013 | 06:45 PM - Posted by netsez (not verified)

OK, You have flooded us with a ton of charts and stats, but can you put in a paragraph or two explaining what the gaming experience FEELS like? Does a game play better with or without SLI/XFIRE, vsync on/off, etc. In the end the gameplay experience is what matters MOST.
That is something hardocp.com does well.

March 30, 2013 | 11:25 PM - Posted by Pendulously (not verified)

"Another stutter metric is going to be needed to catch and quantify them directly."

EXAMPLE: If Average FPS (over 60 seconds) is 100, then Total Frames observed over 60 seconds is 6000.

If ONE SINGLE FRAME is above 100ms, then for the y-axis value '100' (milliseconds), the x-axis value will be '99.9983' (percentile), i.e. one minus (1/6000).

If FOUR FRAMES are above 30ms, then for the y-axis value '30' (milliseconds), the x-axis value will be '99.9933' (percentile), i.e. one minus (4/6000).

If TEN FRAMES are above 20ms, then for the y-axis value '20' (milliseconds), the x-axis value will be '99.8333' (percentile), i.e. one minus (10/6000).

So instead of PERCENTILE on the X-AXIS, you can put NUMBER OF FRAMES on the X-AXIS. For the y-axis value of '100' (ms), the x-axis value will be '1' (frame), for y-axis '30' (ms), the x-axis will be '4' (frames), and so on.

March 31, 2013 | 05:27 AM - Posted by Anonymous (not verified)

same story in every place I go. 2 lines to say 7970 is better than 680 1x1 than 3 pages to say about temperature/power and shutter when it's 2x2 ... =_=

March 31, 2013 | 11:26 AM - Posted by Anonymous (not verified)

Well, These articles will change the graphic card industry a lot

but I also have a suggestion, not just test the graphic card on pc, maybe it can be used to test the console(xbox360, ps3 or xbox720 ps4 in the future) games.

April 1, 2013 | 10:18 AM - Posted by FernanDK (not verified)

hey pcper's guys,

you have my deepest sympathies. awesome job you got there. thank you for sharing and introducing us to this new benchmark system, it is for sure facts more trustful than any other way the majority uses to measure FPS nowadays.

keep it up. you have my support and probably the whole community is also on your side.

April 2, 2013 | 04:13 AM - Posted by Anonymous (not verified)

I find your VSYNC tests to be invalid because the game needs to be able to run at minimum of 60 FPS for VSYNC to work properly, FPS can not be allowed to drop below 60. So you should actually fiddle with the game settings until you get a minimum of 60 FPS and only then enable VSYNC and test the results, because that's what a knowledgeable player would do. Nobody plays the game at full settings and VSYNC enabled if they can't get MIN 60 FPS, that's just stupid.

April 2, 2013 | 08:36 AM - Posted by JC (not verified)

I do. I can't stand the tearing and would rather have a lower frame rate that's free of tearing.

April 3, 2013 | 05:16 AM - Posted by Anonymous (not verified)

Reading is hard, isn't it ?

I said that in order for the VSYNC tests to be valid the game needs to run over 60 FPS constantly, because that's what a knowledgeable user would do in order to enjoy smooth gameplay without tearing.

Now if you're stupid enough to run VSYNC while under 60 FPS then that's your loss, still doesn't mean jack shit for the purposes of me challenging their VSYNC test.

April 2, 2013 | 01:46 PM - Posted by MadManniMan (not verified)

Hey there guys,
when will the missing parts of the article series follow? I am like a cat on a hot tin roof here ...

April 2, 2013 | 04:36 PM - Posted by HalloweenJack (not verified)

FRAPS can manipulate what it receives so if this software is on the same place , then it too can manipulate anything - also , having yet another layer WILL slow down everything , once the frame is finished its being intercepted before being sent to the RAMDAC.

So basically this NVidia software , which you`ve had for a year , has helped NVidia `silently` attempt fix there own SLI frame syncing issues , but now you`ve `come out` against AMD.

NVidia pay roll that good now? they do need help since tegra 5 will be old before its out , given that ARM have now sampled X64 V57 on 16nm

April 3, 2013 | 07:53 AM - Posted by ThorAxe

I'm not sure if you are being serious or are just posting to make us laugh. I hope it's the latter because no one can be that stupid, can they?

The frame is intercepted before being sent to RAMDAC??...LOL!!!

April 18, 2013 | 11:20 PM - Posted by CoderX71 (not verified)

NVIDIA was responsible for developing the color overlay that sits between the game and DirectX (in the same location of the pipeline as FRAPS essentially)

Yeah ... nVidia would do anything untoward here right?

April 18, 2013 | 11:20 PM - Posted by CoderX71 (not verified)

NVIDIA was responsible for developing the color overlay that sits between the game and DirectX (in the same location of the pipeline as FRAPS essentially)

Yeah ... nVidia would do anything untoward here right?

April 4, 2013 | 07:46 AM - Posted by Anonymous (not verified)

Thank you very much for finally exposing this AMD Crossfire scam using runts and ghost frames to artificially inflate frame rate and sell its products based on false data. I applaud you for giving us a clear picture of AMD's dirty runt&ghost frames games, AMD plays many benchmark games that's why AMD fled BAPco in shame. Nothing AMD claims is to believed, and if it wasn't for honest independent websites like yours that expose AMD for what it has become, AMD would still be selling their inferior products based on false viral marketing and propaganda pumping message board bullying.

Being the under dog doesn't give AMD the right to blatantly pump false benchmarks as fact. If AMD has to cheat to sell its products then AMD needs to be exposed as a cheater and people need to be made aware of it, thank you for doing so.

April 5, 2013 | 04:35 AM - Posted by Joseph C. Carbone III (not verified)

Dear Mr. Stroud,

Thank you for a tremendous amount of work, diligence, and integrity...

I thoroughly enjoyed the video with you and Tom Petersem. Although, I have to mention; I have been very disappointed in Nvidia since my purchase of a group of GTX 480s, believing, from day one, I had thrown away more than $1500 for three unmanageable 1500 Watt hairdryers marketed as graphics cards, which were subsequently relabeled GTX 580s, once the bugs were worked out, kind of like Microsoft's Vista to XP, kind of like scamming on people--no, definitely scamming.

I have always been an enthusiast of the Nvidia since the days of 3dfx and had likewise always enjoyed anticipating and buying Nvidia's new products, and the GTX Titan is awesome.

With memories of ATI, Matrox, and Nvidia (I still have my RIVA 128), a home has been found within my memories, and that is why I am excited about what you and the rest of PC Perspective have done and are going to do. With collaboration you-all are moving a beloved industry onward toward a better future for us and for the companies we want to succeed.

Sincerely,
Joseph C. Carbone III; 05 April 2013

April 8, 2013 | 12:19 PM - Posted by Anonymous (not verified)

I find the testing methodology to be flawed.

Using a competitors product where they get to define the capture data and define the test results is not accurate scientific method.

Has anyone asked the question as to why Nvidia defined a runt as 21 scan lines?

Has an analysis been done to see how many scan lines an Nvidia product has produced that are not considered runts because they are above the arbitrary 21 scan lines, vs how many on an AMD product just happen to be at 21 or below?

I am skeptical, as you should be too as reviewers and analyzers, that the 21 scan lines was chosen because, for whatever reason, Nvidia products produce frames that are 22 scan lines or greater, so therefor 21 scan lines were chosen as runts.

This would seem to be even supported by your own data, where Nvidia products do not produce any runts on any test, which would seem a remarkable situation unless you consider the fact that they get to define the metric that determines what a runt is.

Is it possible to rerun these tests where the runt was defined as perhaps 28 scan lines instead? 35 scan lines? Greater? How about report only fully rendered frames as the framerate?

These above would be much more accurate tests, as they would determine if there was a noticeable jump in AMD related runts and what the number of Nvidia related runts would become.

Would it surprise anyone that Nvidia produced 22 scan line runts vs AMD 21 scan line runts and that is why Nvidia chose 21 scan lines as the break point?

April 9, 2013 | 05:50 AM - Posted by Danieldp

Hi guys,

I am new to the forum and have been following this discussion for a while now and thought I would register. I currently own nvidia (690) and I have also owned ATI (7970 CF).

In regards to your comment sir or mam, I think you make an extremely valid point. I have been looking through these articles and every time I see the graphs I almost feel like there is a slight bias, maybe not intentional, as you mentioned, but I do think it is there. looking at some of the CF results I just don't find them accurate in regards to user experience. I remember playing on one 7970 ghz in crysis 3 then upgrading to a CF setup and I could immediately tell the difference the game play was a lot better, there was the occasional stuttering but nothing as bad as these graphs make them out to be... This is a slight subjective view but I do think we need testing methods that are more thorough and as you say, more towards the scientific method.

I own nvidia and I think anyone who does will get exited at what these graphs are saying as for amd owners you will probably be a bit disheartened, from everything I have read I can only conclude that nvidia has had a big hand in this testing {mostly indirect) and therefore I can not take these results too serious.

There is definitely truth in what these graphs say but I think it is blown a little out of proportion, especially with the parameters of testing set to be nvidia optimized.

I welcome critics but please be professional about it.

Thanks,
Dan

April 9, 2013 | 09:53 AM - Posted by Danieldp

Hi guys,

This confirms what I have been saying the CROSSFIRE SETUP WINS against the TITAN in FRAME TIME VARIANCE!

http://www.tomshardware.com/reviews/crysis-3-performance-benchmark-gamin...

Still lolve my 690 though =)

April 9, 2013 | 08:41 PM - Posted by Danieldp

And by win I mean JUST. But, it does, the results are different from PCP. Check the triple monitor frame time deviance as well.

http://www.tomshardware.com/reviews/crysis-3-performance-benchmark-gamin...

The important thing here is to note that ALL setups are BELLOW 20 ms... what does that say? All are playable.

so in my opinion, from what I just saw, if I have to choose between a titan or two sapphire vapor-x 6gig in CF, I would choose the latter. Unless you want to SLI with the titan, then it is titan FTW! But, that is just what I would do based on the results.

Again, what is shocking to me is how different these results are to PCP... Toms hardware has been around longer than any site I know of and they have always been held in the highest regard (not an opinion).

It will be interesting to see what the new AMD runt fix in July will do to these results. I am thinking they might be more than just back on the board after that (speculation).

Interesting, is it not?

April 12, 2013 | 04:13 AM - Posted by JCCIII

Dear Daniel,

With Tom’s Hardware disclosing, from the link you provided, that they do not consider their dual-card and dual-GPU results as accurate and, with their tests showing the Radeon HD 7950’s time variance at 23.8 in the 95th percentile, how is it you see PC Perspective’s research as invalidated?

I first went to Maximum PC and Tom’s Hardware to try to understand this complicated subject, but they are both behind on the subject of why we are spending much money and getting unpredictable quality on our screens.

Ultimately, good research is fundamental, and it is obvious PC Perspective has committed a great deal of resources in order to be helpful and do a good job. They have worked to present a balanced perspective for us to consider, and it is still being translated into tangible empirical value. However, I am certain, with collaboration of Tom’s Hardware and others, PC Perspective’s results are accurate, significant, and will benefit us all.

Sincerely, Joseph C. Carbone III; 12 April 2013

April 12, 2013 | 04:15 AM - Posted by JCCIII

...

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.