What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

AMD 13.8 Frame Pacing Driver Tested & Benchmarked

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Earlier this year, AMD launched their eagerly anticipated HD 7990. Boasting a pair of Tahiti XT cores and 6GB of memory, it had the potential to unseat NVIDIA’s GTX 690 as the king of the graphics card world. Unfortunately, that didn’t happen. Instead of powering itself to an enviable position, the HD 7990 was saddled with stuttering frame delivery which not only ruined its in-game experience but also stabbed performance through the heart. A fix was needed, and fast.

The stuttering which is painfully evident on all of AMD’s dual card solutions is a byproduct of inter-chip communications which strive to deliver frames at exactly the pace a game engine asks for them. Basically, it wants them as quickly as possible regardless of the consequences. Without going into the gory details, let’s just say this overly rapid frame delivery wreaks havoc on in-game fluidity and image quality as some frames are dropped, some take longer to render and others are only partially rendered. There’s just no rhyme or reason unless an outside factor steps in to smooth things out. The only way around this was to either enable triple buffering (which only worked in some applications) or apply V-Sync which causes its own set of limitations.

FRAME-PACING-19.jpg

While NVIDIA has been actively addressing these frame pacing issues for years now, AMD is only getting around to fixing them now. Luckily for the folks on AMD’s engineering teams, the stuttering wasn’t an issue with their underlying architecture. Rather, the culprit was software-based, rooted deeply within the driver stack. They’ve been promising a solution to stuttering for months and months but nothing really happened. We saw glimpses of what a so-called “frame pacing” driver would look like but an actual public release proved to be elusive…until now that is.

With the 13.8 beta driver AMD is instituting what they call Frame Pacing. This technology performs an on-the-fly, frame by frame analysis so instead of rendering based on a game engine’s requests, tries to smooth things out and deliver consistency. Contrary to popular belief, this won’t negatively impact raw framerates since AMD is simply endeavoring to eliminate dropped, partially rendered and high latency frames. If this is accomplished, the number of frames that actually make their way onto the screen should, in theory, increase.

FRAME-PACING-18.png

Frame Pacing is automatically installed with the 13.8 driver suite (available today on AMD’s driver download portal) but, in the event of issues AMD allows it to be turned on and off. In our experience, the system will need a reboot to ensure a setting change is actually implemented but otherwise, the toggle worked perfectly. According to AMD there aren’t any side effects either and they claim that input lag hasn’t been affected.

Compatibility isn’t an issue either since the new driver won’t be restricted to just the GCN architecture. Users of older cards like the HD 6000-series will get some Frame Pacing lovin’ too.

As with all good things in life, AMD’s Frame Pacing does have some limitations. It will only be implemented on DX10 and DX11 applications and multi monitor support isn’t being rolled out yet. That will be cold comfort for the thousands of Radeon users still making their way through The Witcher 2 or the many other stutter-filled yet popular DX9 titles. Eyefinity and 4K users will also be left without support for the time being but luckily, AMD isn't trying to sweep this under the rug.

So how does AMD’s new Frame Pacing perform? With a HD 7990 in hand we wanted to find out.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Test System & Setup

Main Test System

Processor: Intel i7 3930K @ 4.5GHz
Memory: Corsair Vengeance 32GB @ 1866MHz
Motherboard: ASUS P9X79 WS
Cooling: Corsair H80
SSD: 2x Corsair Performance Pro 256GB
Power Supply: Corsair AX1200
Monitor: Samsung 305T / 3x Acer 235Hz
OS: Windows 7 Ultimate N x64 SP1



Drivers:
AMD 13.8 BETA


*Notes:

- All games tested have been patched to their latest version

- The OS has had all the latest hotfixes and updates installed

- All scores you see are the averages after 3 benchmark runs

All IQ settings were adjusted in-game and all GPU control panels were set to use application settings
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Assassin’s Creed III / Crysis 3

NOTE: All tests are done with FCAT. FRAPS is no longer being used.

Assassin’s Creed III (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/RvFXKwDCpBI?rel=0" frameborder="0" allowfullscreen></iframe>​

The third iteration of the Assassin’s Creed franchise is the first to make extensive use of DX11 graphics technology. In this benchmark sequence, we proceed through a run-through of the Boston area which features plenty of NPCs, distant views and high levels of detail.


2560 x 1440

FRAME-PACING-1.jpg



Crysis 3 (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/zENXVbmroNo?rel=0" frameborder="0" allowfullscreen></iframe>​

Simply put, Crysis 3 is one of the best looking PC games of all time and it demands a heavy system investment before even trying to enable higher detail settings. Our benchmark sequence for this one replicates a typical gameplay condition within the New York dome and consists of a run-through interspersed with a few explosions for good measure Due to the hefty system resource needs of this game, post-process FXAA was used in the place of MSAA.


2560 x 1440

FRAME-PACING-2.jpg


The first tests show some real promise for this new driver since both Crysis 3 and Assassin's Creed III show impressive framerate improvements. There aren't any noticeable performance drop-offs which is impressive since both runs use the exact same driver.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Dirt: Showdown / Far Cry 3

NOTE: All tests are done with FCAT. FRAPS is no longer being used.

Dirt: Showdown (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/IFeuOhk14h0?rel=0" frameborder="0" allowfullscreen></iframe>​

Among racing games, Dirt: Showdown is somewhat unique since it deals with demolition-derby type racing where the player is actually rewarded for wrecking other cars. It is also one of the many titles which falls under the Gaming Evolved umbrella so the development team has worked hard with AMD to implement DX11 features. In this case, we set up a custom 1-lap circuit using the in-game benchmark tool within the Nevada level.


2560 x 1440

FRAME-PACING-3.jpg



Far Cry 3 (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/mGvwWHzn6qY?rel=0" frameborder="0" allowfullscreen></iframe>​

One of the best looking games in recent memory, Far Cry 3 has the capability to bring even the fastest systems to their knees. Its use of nearly the entire repertoire of DX11’s tricks may come at a high cost but with the proper GPU, the visuals will be absolutely stunning.

To benchmark Far Cry 3, we used a typical run-through which includes several in-game environments such as a jungle, in-vehicle and in-town areas.



2560 x 1440

FRAME-PACING-4.jpg


Here we are seeing some interesting results with the new frame pacing feature lacking any differentiation between the on and off settings. As a matter of fact, frame pacing actually decreases performance in Dirt Showdown.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Hitman Absolution / Max Payne 3

NOTE: All tests are done with FCAT. FRAPS is no longer being used.

Hitman Absolution (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/8UXx0gbkUl0?rel=0" frameborder="0" allowfullscreen></iframe>​

Hitman is arguably one of the most popular FPS (first person “sneaking”) franchises around and this time around Agent 47 goes rogue so mayhem soon follows. Our benchmark sequence is taken from the beginning of the Terminus level which is one of the most graphically-intensive areas of the entire game. It features an environment virtually bathed in rain and puddles making for numerous reflections and complicated lighting effects.


2560 x 1440

FRAME-PACING-5.jpg



Max Payne 3 (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/ZdiYTGHhG-k?rel=0" frameborder="0" allowfullscreen></iframe>​

When Rockstar released Max Payne 3, it quickly became known as a resource hog and that isn’t surprising considering its top-shelf graphics quality. This benchmark sequence is taken from Chapter 2, Scene 14 and includes a run-through of a rooftop level featuring expansive views. Due to its random nature, combat is kept to a minimum so as to not overly impact the final result.


2560 x 1440

FRAME-PACING-6.jpg



Unfortunately, Hitman shows that AMD may not have optimized frame pacing all that well but as we will see in the Frame Time sections, there's a method behind this madness. Max Payne meanwhile doesn't show any differentiation between the two settings.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Metro: Last Light / Tomb Raider

NOTE: All tests are done with FCAT. FRAPS is no longer being used.

Metro: Last Light (DX11)


<iframe width="640" height="360" src="http://www.youtube.com/embed/40Rip9szroU" frameborder="0" allowfullscreen></iframe>​

The latest iteration of the Metro franchise once again sets high water marks for graphics fidelity and making use of advanced DX11 features. In this benchmark, we use the Torchling level which represents a scene you’ll be intimately familiar with after playing this game: a murky sewer underground.


2560 x 1440

FRAME-PACING-7.jpg



Tomb Raider (DX11)


<iframe width="560" height="315" src="http://www.youtube.com/embed/okFRgtsbPWE" frameborder="0" allowfullscreen></iframe>​

Tomb Raider is one of the most iconic brands in PC gaming and this iteration brings Lara Croft back in DX11 glory. This happens to not only be one of the most popular games around but it is also one of the best looking by using the entire bag of DX11 tricks to properly deliver an atmospheric gaming experience.

In this run-through we use a section of the Shanty Town level. While it may not represent the caves, tunnels and tombs of many other levels, it is one of the most demanding sequences in Tomb Raider.


2560 x 1440

FRAME-PACING-8.jpg


Here we truly begin to see the benefits of frame pacing. There are less significant framerate drops and more consistent overall performance.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Onscreen Frame Times w/FCAT

Onscreen Frame Times w/FCAT


When capturing output frames in real-time, there are a number of eccentricities which wouldn’t normally be picked up by FRAPS but are nonetheless important to take into account. For example, some graphics solutions can either partially display a frame or drop it altogether. While both situations may sound horrible, these so-called “runts” and dropped frames will be completely invisible to someone sitting in front of a monitor. However, since these are counted by its software as full frames, FRAPS tends to factor them into the equation nonetheless, potentially giving results that don’t reflect what’s actually being displayed.

With certain frame types being non-threatening to the overall gaming experience, we’re presented with a simple question: should the fine-grain details of these invisible runts and dropped frames be displayed outright or should we show a more realistic representation of what you’ll see on the screen? Since Hardware Canucks is striving to evaluate cards based upon and end-user experience rather than from a purely scientific standpoint, we decided on the latter of these two methods.

With this in mind, we’ve used the FCAT tools to add the timing of partially rendered frames to the latency of successive frames. Dropped frames meanwhile are ignored as their value is zero. This provides a more realistic snapshot of visible fluidity.


FRAME-PACING-10.jpg

FRAME-PACING-11.jpg

FRAME-PACING-12.jpg

FRAME-PACING-13.jpg


As we move onto frame times, it seems like the frame pacing really does make a difference. Again and again it was able to minimize stuttering by significantly lowering frame times while minimizing latency deltas. However, statistical charts only tell a fraction of the story. The HD 7990 with frame pacing enabled feels consistently smoother in every game, even Far Cry 3 which has posed some considerable problems for AMD dual GPU setups in the past.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Onscreen Frame Times w/FCAT (pg.2)

Onscreen Frame Times w/FCAT (pg.2)


When capturing output frames in real-time, there are a number of eccentricities which wouldn’t normally be picked up by FRAPS but are nonetheless important to take into account. For example, some graphics solutions can either partially display a frame or drop it altogether. While both situations may sound horrible, these so-called “runts” and dropped frames will be completely invisible to someone sitting in front of a monitor. However, since these are counted by its software as full frames, FRAPS tends to factor them into the equation nonetheless, potentially giving results that don’t reflect what’s actually being displayed.

With certain frame types being non-threatening to the overall gaming experience, we’re presented with a simple question: should the fine-grain details of these invisible runts and dropped frames be displayed outright or should we show a more realistic representation of what you’ll see on the screen? Since Hardware Canucks is striving to evaluate cards based upon and end-user experience rather than from a purely scientific standpoint, we decided on the latter of these two methods. With this in mind, we’ve used the FCAT tools to add the timing of runted to the latency of successive frames. Dropped frames meanwhile are ignored as their value is zero. This provides a more realistic snapshot of visible fluidity.


FRAME-PACING-14.jpg

FRAME-PACING-15.jpg

FRAME-PACING-16.jpg

FRAME-PACING-17.jpg


The hits continue here with frame pacing once again proving its worth in every game but Metro Last Light where both results are very much even. Tomb Raider in particular seems some impressive benefits from AMD's new technology.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Conclusion; Better Late Than Never

Conclusion; Better Late Than Never


After what seems like endless waiting and more than a few rumors about its impending release, AMD’s Frame Pacing driver has finally arrived and it works extremely well. For the most part, it delivered a notable difference for onscreen frame delivery without negatively impacting other aspects of gameplay.

Frame pacing performed up to expectations and then some. Its impact on frame times was, in many instances, extreme, with a significant reduction in latencies and a virtual elimination of dropped and runted frames. Granted, things weren't always perfect and there were still some evident hiccups but that’s to be expected with beta software.

Looking outside of our little statistical FCAT bubble, the difference between disabling and enabling frame pacing was dramatic. Nearly every game we tried was drastically smoother, without many of the telltale stutters which typically occurred when using two AMD GPUs in Crossfire. Some stuttering still occurs in certain games but for the most part, it has been reduced to a manageable level. Even Far Cry 3, a perpetual problem child, showed some notable improvements. This is an impressive achievement considering where Crossfire solutions were before 13.8’s release.

FRAME-PACING-9.jpg

The actual framerate performance of this driver will –at least when frame pacing is enabled- likely be a hotly debated subject. In some instances, framerates actually drop by a good amount but that was something AMD actively planned for. This situation does raise a very good question though: what makes a graphics card "fast"? With frame times and onscreen fluidity taken into consideration, it's obviously not enough to have a GPU that can pump out ridiculous framerates without giving any thought to the actual gaming experience.

With this in mind, AMD's frame pacing works in a number of different ways. It smoothes out the large latency deltas which used to occur from one frame to the next and it does so without sinking average frames per second into unplayable territory. This delicate balancing act has been accomplished without negatively affecting in-game, perceptual performance. Remember; a dual GPU solution working at 100FPS with loads of stutter will always "feel" slower than one operating at 80FPS without any noticeable frame delivery issues. A refined gameplay experience rather than a raw, uncontrolled torrent of frames is what AMD is banking on here and that’s something we can appreciate

We should also mention that frame pacing is still a work in progress and it isn’t ready for anything beyond the beta stage just yet. Optimizations are still being worked on and performance should gradually increase. No, that isn't likely to be something you want to hear right now, after waiting for these drivers for months on end but this is all we've got for now.

There have been some theories that AMD’s attempt to reign in stuttering would result in a noticeable increase in input lag, similar to what happens with V-Sync enabled. In our experience, that didn’t happen. When using a corded mouse, input lag wasn’t adversely affected, nor were there any secondary issues like artifacts or rendering errors.

Frame pacing represents an excellent bonus for current Crossfire users, but what about the HD 7990? Does smoother gameplay make it any more appealing? The answer to that is multi-faceted. The HD 7990 crashed and burned for a number of reasons; it was late, power hungry, loud, featured epic amounts of coil whine and it highlighted AMD’s frame timing issues. Performance has now been honed into something resembling a bona fide gaming solution but at this point, months after its release, we feel this ship has already sailed. Anyone who bought a HD 7990 will be breathing a sigh of relief though.

AMD’s new driver is an impressive step forward but we can’t help feeling that it took AMD ages to get to this point. While review sites have been actively reporting and recording AMD’s problems with frame times for the past year or so, the root causes have been around for far longer. Remember, NVIDIA stepped in and fixed their issues without any fanfare years ago.

Even though frame pacing does represent a large improvement for Crossfire solutions, we just can’t bring ourselves to laud it as a technological achievement. Late is certainly better than never but with the 13.8 beta drivers, AMD is simply fixing something which shouldn’t have been broken for so long in the first place.
 
Last edited:
Top