• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Catalyst "Frame Pacing" Driver Results Thread

erocker

Senior Moderator
Staff member
Joined
Jul 19, 2006
Messages
42,445 (10.11/day)
Likes
18,110
#1
So.. New drivers incoming today or tomorrow. They will supposedly be named 13.8's... "8" being August. Anyways, there are two programs which you can use to find out what your frame latency is in games.

The first tool you will need is FRAPS. Get it here: http://www.fraps.com/download.php

The second tool you need is FRAFS Bench Viewer. Get it here: http://sourceforge.net/projects/frafsbenchview/

To begin, open FRAPS and tick the "FPS" tab. Under Benchmark Settings, the only box that needs to be checked is "Frametimes" You can set the duration of the test to a number of your choosing.



After you have recorded the frametimes, you then simply open up FRAFS and drag the .csv file that FRAPS created into FRAFS.



I'm going to start benching some games in CrossFire using the current 13.6 Beta 2 drivers now. I will post up the new drivers once they are available.
 

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#2
Great, except this won't cover issues that aren't seen here, that appear over the cable. AMD agreed that FRAPS wasn't showing the true story about this issue already, because of where in the render pipeline it accesses the data from, after which a lot of stuff can still happen.


That said, I'm still interested to see if there is an impact at this level as well, too. It's all pretty useful info, since AMD isn't the only one with issues at times. Oh and look, jUly 31st..no driver yet, and it's 1:30 PM MST. Boo AMD, BOO...lulz.
 
Joined
Apr 8, 2012
Messages
575 (0.27/day)
Likes
188
System Name Main rig
Processor Intel i7-4790K Devils Canyon
Motherboard Asus Z97-A
Cooling Antec Kuhler 620
Memory 16GB Corsair Vengence Pro
Video Card(s) Sapphire R9-290 Vapor-X
Storage Samsung 840 EVO 1TB
Case Corsair 600T
Audio Device(s) Sound Blaster X-FI HD Platinum
Power Supply Corsair AX750 Gold
#3
it should show some changes as the pacing should back up the fill buffers, which is what fraps measures.
 

erocker

Senior Moderator
Staff member
Joined
Jul 19, 2006
Messages
42,445 (10.11/day)
Likes
18,110
#4
Great, except this won't cover issues that aren't seen here, that appear over the cable. AMD agreed that FRAPS wasn't showing the true story about this issue already, because of where in the render pipeline it accesses the data from, after which a lot of stuff can still happen.
When dealing with what FPS Fraps gives. We're not dealing with that. We are taking the frametimes and putting it into FRAFS. The results are accurate and it does cover the issue.
 

MxPhenom 216

Corsair Fanboy
Joined
Aug 31, 2010
Messages
12,147 (4.50/day)
Likes
3,708
Location
Seattle, WA
System Name The Battlestation
Processor Intel Core i7 4770k @ 4.2GHZ 1.275v
Motherboard MSi Z97 Gaming 5
Cooling EK Supremacy w/ EK Coolstream PE360
Memory G. Skill Trident X 16Gb (4x4GB) 2400mhz @ 1.65v
Video Card(s) MSi GTX1070 Gaming X 8GB @ 2GHz
Storage Samsung 830 128GB SSD, Crucial MX200 500GB, Seagate Barracuda 2TB (2x 1TB Partitions)
Display(s) Qnix QX2710 27" 2560 x 1440 PLS @ 100hz
Case Phantek Enthoo Evolv ATX TG
Audio Device(s) MSi Gaming AudioBoost ALC1150 w/ Sennheiser Game Ones
Power Supply Seasonic Flagship Prime Platinum 850
Mouse Steelseries Rival 310 w/ QCK Mass mat
Keyboard Corsair K70 w/ MX Browns and Red Backlit
Software Windows 10 Pro 64-Bit
Benchmark Scores Firestrike: 15439
#5
Great, except this won't cover issues that aren't seen here, that appear over the cable. AMD agreed that FRAPS wasn't showing the true story about this issue already, because of where in the render pipeline it accesses the data from, after which a lot of stuff can still happen.


That said, I'm still interested to see if there is an impact at this level as well, too. It's all pretty useful info, since AMD isn't the only one with issues at times. Oh and look, jUly 31st..no driver yet, and it's 1:30 PM MST. Boo AMD, BOO...lulz.
LOL :laugh:

I was watching some videos, mainly of Crysis 3 from PCPer and One side was driver prior to the Prototype they were given, and then right was with the Prototype and it was like a night and day different. They then dropped the speed of the video to 50% and I was amazed. Left side was super stuttery, right side was butter smooth!
 

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#6
When dealing with what FPS Fraps gives. We're not dealing with that. We are taking the frametimes and putting it into FRAFS. The results are accurate and it does cover the issue.
FRAFS reads the timestamps put on the frame by FRAPS. But rather than showing FPS, it shows the difference beween the render times at that specific point in the pipeline.

The problem here is not in using FRAPS to measure average framerates over the run of a benchmark, but rather when it comes to using FRAPS to measure individual frames. FRAPS is at the very start of the rendering pipeline; it’s before the GPU, it’s before the drivers, it’s even before Direct3D and the context queue. As such FRAPS can tell you all about what goes into the rendering pipeline, but FRAPS cannot tell you what comes out of the rendering pipeline.

So to use FRAPS in this method as a way of measuring frame intervals is problematic. Considering in particular that the application can only pass off a new frame when the context queue is ready for it, what FRAPS is actually measuring is the very start of the rendering pipeline, which not unlike a true pipe is limited by what comes after it. If the pipeline is backed up for whatever reason (context queue, drivers, etc), then FRAPS is essentially reporting on what the pipeline is doing, and not the frame interval on the final displayed frames. Simply put, FRAPS cannot tell you the frame interval at the end of the pipeline, it can only infer it from what it’s seeing.
http://www.anandtech.com/show/6857/amd-stuttering-issues-driver-roadmap-fraps/4
 
Joined
Mar 10, 2010
Messages
5,108 (1.78/day)
Likes
1,653
Location
Manchester uk
System Name Quad GT evo V
Processor FX8350 @ 4.8ghz1.525c NB2.64ghz Ht2.84ghz
Motherboard Gigabyte 990X Gaming
Cooling 360EK extreme 360Tt rad all push/pull, cpu,NB/Vrm blocks all EK
Memory Corsair vengeance 32Gb @1333 cas9
Video Card(s) Rx vega 64 waterblockedEK + Rx580 waterblockedEK
Storage samsung 840(250), WD 1Tb+2Tb +3Tbgrn 1tb hybrid
Display(s) Samsung uea28"850R 4k freesync, samsung 40" 1080p
Case Custom(modded) thermaltake Kandalf
Audio Device(s) Xfi creative 7.1 on board ,Yamaha dts av setup
Power Supply corsair 1000Rmx
Mouse CM optane
Keyboard CM optane
Software Win 10 Pro
Benchmark Scores 15.69K best overall sandra so far
#7
My god they are making us wait for these wtf amd already.
Im actually going to try this this time erocker ,benches only though.
Im probably day dreaming anyway since its unlikely to be applicable to my fives (58##).
 

erocker

Senior Moderator
Staff member
Joined
Jul 19, 2006
Messages
42,445 (10.11/day)
Likes
18,110
#8
FRAFS reads the timestamps put on the frame by FRAPS. But rather than showing FPS, it shows the difference beween the render times at that specific point in the pipeline.



http://www.anandtech.com/show/6857/amd-stuttering-issues-driver-roadmap-fraps/4
If that was the case wouldn't the results be identical using similar systems with different GPU's? Either way, it's nothing official... For fun, if you will. I do know that using this method one can clearly see that Nvidia has the advantage in terms of frame pacing regardless of where FRAPS gets it's information from.
 

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#9
If that was the case wouldn't the results be identical using similar systems with different GPU's?

No, but rather than try to explain it all(since it's pretty damn complicated and I'm not sure I exactly understand all of it myself), I just suggest reading the Anandtech article I linked. Both Nvidia and AMD are not exactly happy with FRAPS, because of how it works, but in the end, for FPS testing, it works fine since it still tells you how many frames are rendered, since one cannot be finished before the next, so to speak. It just doesn't tell you how many were actually displayed, and that was the whole point of PC PERSPECTIVE doing all that testing using a second PC, etc...FRAPS clearly didn't show the entire picture, at all.

I do know that using this method one can clearly see that Nvidia has the advantage in terms of frame pacing regardless of where FRAPS gets it's information from.
NVidia has their own issues with multi-GPU as well. There were quite a few times that the NVidia guy in Ryan's video interview had an embarrassed look on his face when they compared the GTX690 results.


Either way, it's just another aspect of the pipeline itself, so all the data is relevant. My only point is that it's not the full picture in and of itself....that requires a completely separate PC and some pretty expensive video capture hardware.
 
Last edited:

erocker

Senior Moderator
Staff member
Joined
Jul 19, 2006
Messages
42,445 (10.11/day)
Likes
18,110
#10
Agreed. However, this bench (for fun/testing purposes) will/should show some improvement to give somewhat of an idea of how things are progressing.
 

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#11
Agreed. However, this bench (for fun/testing purposes) will/should show some improvement to give somewhat of an idea of how things are progressing.
Sure, since it is also possible that things that look good at this stage, may also have a negative effect at the end of render pipe. I dunno, time will tell, and I expect that Ryan @ PCPer will have some data when the driver comes out. Unless they paid him to shut up. :laugh:

Besides, all this conjecture is useless without the driver itself. If I can play BF3 for more than 5 minutes, and not get sick to my stomach, AMD's done a decent job, and that's all I care about. :roll:
 

manofthem

WCG-TPU Team All-Star!
Joined
Jan 9, 2011
Messages
10,728 (4.18/day)
Likes
26,195
Location
Florida
System Name bBox | tOffice | mDonk
Processor i7 4770k @4.3 | i7 2600k @3.5 | i7 4790 @3.8
Motherboard MSI Z87 G45 | MSI Z68 GD55 | MSI H97M G43
Cooling Water | Reeven Justice RC-1204 | can't remember
Memory G Skill TridentX 8GB 2400 | Patriot Viper 3 8GB 1600 | G Skill Ripjaws 8GB 1600
Video Card(s) 2x R9 290's | 980 SC ACX2.0 | n/a
Storage 850 Evo 250GB & 2TB Desktar 5K3000 | 830 256GB & 1.5TB & 3TB Barracudas| 830 256GB
Display(s) QNIX WX2710 (96Hz OC) & Shimian Achieva QH270-Lite | Acer 27HL | n/a
Case 900D | Fractal Define S | n/a
Audio Device(s) Siberia v2 | onboard | n/a
Power Supply Corsair AX1200 | XFX 550Pro | SeaSonic 660XP2
Mouse Razer Mamba 2012 | Corsair M95 | n/a
Keyboard Das Keyboard Pro 4 (blues) | Cougar Attack X3 (blues) | n/a
Software Win 10 Pro x64 | Win 10 Pro x64 | Win 10 Pro x64
#12
Well, FRAPS tells me my fps sucks in Far Cry 3, so I'm looking forward to some better numbers with this new driver release, hopefully.
 

T4C Fantasy

CPU & GPU DB Maintainer
Staff member
Joined
May 7, 2012
Messages
1,239 (0.60/day)
Likes
609
Location
Rhode Island
#13
is it normal im getting 29 fps in tomb raider with a hd 7970 max settings? it doesn't seem like its lagging at all but fraps says 29/30/31 and the 7970 is clocked at 1125/1575, single gpu config on 1080p

13.150.100.1000 8.1 Preview V2 drivers
Time


FPS


Rank
 
Last edited:

MxPhenom 216

Corsair Fanboy
Joined
Aug 31, 2010
Messages
12,147 (4.50/day)
Likes
3,708
Location
Seattle, WA
System Name The Battlestation
Processor Intel Core i7 4770k @ 4.2GHZ 1.275v
Motherboard MSi Z97 Gaming 5
Cooling EK Supremacy w/ EK Coolstream PE360
Memory G. Skill Trident X 16Gb (4x4GB) 2400mhz @ 1.65v
Video Card(s) MSi GTX1070 Gaming X 8GB @ 2GHz
Storage Samsung 830 128GB SSD, Crucial MX200 500GB, Seagate Barracuda 2TB (2x 1TB Partitions)
Display(s) Qnix QX2710 27" 2560 x 1440 PLS @ 100hz
Case Phantek Enthoo Evolv ATX TG
Audio Device(s) MSi Gaming AudioBoost ALC1150 w/ Sennheiser Game Ones
Power Supply Seasonic Flagship Prime Platinum 850
Mouse Steelseries Rival 310 w/ QCK Mass mat
Keyboard Corsair K70 w/ MX Browns and Red Backlit
Software Windows 10 Pro 64-Bit
Benchmark Scores Firestrike: 15439
#14
is it normal im getting 29 fps in tomb raider with a hd 7970 max settings? it doesn't seem like its lagging at all but fraps says 29/30/31 and the 7970 is clocked at 1125/1575, single gpu config on 1080p

13.150.100.1000 8.1 Preview V2 drivers
Time
http://img.techpowerup.org/130731/TombRaider2013-07-3116-14-29-59-Time.png

FPS
http://img.techpowerup.org/130731/TombRaider2013-07-3116-14-29-59-FPS.png

Rank
http://img.techpowerup.org/130731/TombRaider2013-07-3116-14-29-59-Rank.png
You should be getting about 50-60FPS lol.
 
Joined
Jan 21, 2013
Messages
3,800 (2.09/day)
Likes
1,320
System Name Out of the game..
Benchmark Scores d1nky@hwbot.org
#15
^^^ i thought this frametime/microstutter/frame latency was for dual card configs??

and ive been running two cards for a few days and played all sorts of games and havent noticed it once.
 
Joined
Apr 8, 2012
Messages
575 (0.27/day)
Likes
188
System Name Main rig
Processor Intel i7-4790K Devils Canyon
Motherboard Asus Z97-A
Cooling Antec Kuhler 620
Memory 16GB Corsair Vengence Pro
Video Card(s) Sapphire R9-290 Vapor-X
Storage Samsung 840 EVO 1TB
Case Corsair 600T
Audio Device(s) Sound Blaster X-FI HD Platinum
Power Supply Corsair AX750 Gold
#17

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#19
and ive been running two cards for a few days and played all sorts of games and havent noticed it once.
I still see issues at time, and the issues generate an actual physical reaction for me, like motion sickness. I don't care about graphs and what-not.

Anyway, simply put, many users don't see any issues, but it could be construed that those that see no issues are simply not "sensitive" enough (ie. perceptually challenged) to see the actual problem. Usually these sorts of conversations devolve into discussing the ins and outs of the problem, but that's not needed, really. It's a real problem, so much so that AMD has spent months-worth of man-hours, if not years, trying to come up with a suitable fix. IF you don't see the problem, that's great, enjoy your Dual GPUs.

Let the nerdrage commence.

Anyone actually expecting AMD to deliver on-time, when they nearly never do when it comes to drivers and this situation, is simply delusional.:laugh:
 
Last edited:

T4C Fantasy

CPU & GPU DB Maintainer
Staff member
Joined
May 7, 2012
Messages
1,239 (0.60/day)
Likes
609
Location
Rhode Island
#20
I still see issues at time, and the issues generate an actual physical reaction for me, like motion sickness. I don't acre about graphs and what-not.

Anyway, simply put, many users don't see any issues, but it could be construed that those that see no issues are simply not "sensitive" enough (ie. mental challenged) to see the actual problem. Usually these sorts of conversations devolve into discussing the ins and outs of the problem, but that's not needed, really. It's a real problem, so much so that AMD has spent months-worth of man-hours, if not years, trying to come up with a suitable fix. IF you don't see the problem, that's great, enjoy your Dual GPUs.
they said they fixed the single gpu config, the latency from my pics are ridiculous for single gpu are they not?
 
Joined
Jan 21, 2013
Messages
3,800 (2.09/day)
Likes
1,320
System Name Out of the game..
Benchmark Scores d1nky@hwbot.org
#21
LOL ^^^

im waiting for the trolls to jump on this, being late thing.

@ dave theres loads of people with 2 cards that dont notice it, maybe its the over sensitive types that do
 

MxPhenom 216

Corsair Fanboy
Joined
Aug 31, 2010
Messages
12,147 (4.50/day)
Likes
3,708
Location
Seattle, WA
System Name The Battlestation
Processor Intel Core i7 4770k @ 4.2GHZ 1.275v
Motherboard MSi Z97 Gaming 5
Cooling EK Supremacy w/ EK Coolstream PE360
Memory G. Skill Trident X 16Gb (4x4GB) 2400mhz @ 1.65v
Video Card(s) MSi GTX1070 Gaming X 8GB @ 2GHz
Storage Samsung 830 128GB SSD, Crucial MX200 500GB, Seagate Barracuda 2TB (2x 1TB Partitions)
Display(s) Qnix QX2710 27" 2560 x 1440 PLS @ 100hz
Case Phantek Enthoo Evolv ATX TG
Audio Device(s) MSi Gaming AudioBoost ALC1150 w/ Sennheiser Game Ones
Power Supply Seasonic Flagship Prime Platinum 850
Mouse Steelseries Rival 310 w/ QCK Mass mat
Keyboard Corsair K70 w/ MX Browns and Red Backlit
Software Windows 10 Pro 64-Bit
Benchmark Scores Firestrike: 15439
#22
Tress FX has a huge hit on FPS for a little visual pron in TR
It does, but 7970s are still getting about 50FPS with Tress on. Without it, there performance goes up near 70. TressFX is a much bigger hit on Nvidia cards due to the fact their compute performance is very low, compared to AMD Tahiti cards, which TressFX is a compute based shader.
 
Joined
Apr 21, 2010
Messages
4,665 (1.65/day)
Likes
2,017
#23
I still see issues at time, and the issues generate an actual physical reaction for me, like motion sickness. I don't acre about graphs and what-not.

Anyway, simply put, many users don't see any issues, but it could be construed that those that see no issues are simply not "sensitive" enough (ie. mental challenged) to see the actual problem. Usually these sorts of conversations devolve into discussing the ins and outs of the problem, but that's not needed, really. It's a real problem, so much so that AMD has spent months-worth of man-hours, if not years, trying to come up with a suitable fix. IF you don't see the problem, that's great, enjoy your Dual GPUs.
So your retarded if you don't notice frame latency issues with AMD? (I'm sorry if I'm twisting your words but saying mental(ly) challenged people don't see a difference, what are you smoking Dave?) I could say the same for people who overly focus on the whole frame latency (non) issue whilst MANY people with dual AMD GPU setups for years have not seen any noticeable/real world issues (Fraps/Ffaps and faps aside)

Not meaning to be a dick and apologies if that's not how you meant to come across though you really should read what you post before hand as it could be construed in a way that's offensive, and IF you can't see that, maybe it's your problem bro.
 

cadaveca

My name is Dave
Staff member
Joined
Apr 10, 2006
Messages
16,576 (3.85/day)
Likes
10,943
Location
Parkland County, Alberta
#24
LOL ^^^

im waiting for the trolls to jump on this, being late thing.

@ dave theres loads of people with 2 cards that dont notice it, maybe its the over sensitive types that do
PCPer's testing proved that the FPS measurements given by AMD cards, and what was actually displayed, were not one and the same. So if FPS says 80 FPS, you might only have 40 FPS, or 60, or anything between 0 and 80.

So no...24 FPS works fine for many users for blu-rays, but when it comes to twitch-based FPS titles, the differences between even single GPU to dual-GPU, and pretty glaring. AMD's done a lot of work to alleviate the problem as much as possible, but due to responses form suers that say there is no problem, no AMD is forced to basically develop TWO drivers at the same time...one for performance, and one for quality. If this issue was just people being "overly sensitive", they'd not go to that expense, really.

So your retarded if you don't notice frame latency issues with AMD? (I'm sorry if I'm twisting your words but saying mental(ly) challenged people don't see a difference, what are you smoking Dave?) I could say the same for people who overly focus on the whole frame latency (non) issue whilst MANY people with dual AMD GPU setups for years have not seen any noticeable/real world issues (Fraps/Ffaps and faps aside)

Not meaning to be a dick and apologies if that's not how you meant to come across though you really should read what you post before hand as it could be construed in a way that's offensive, and IF you can't see that, maybe it's your problem bro.
meh. And yeah, I knew EXACTLY what I was posting and how it would be interpreted. You should expect that of my posts ALWAYS. ;) I just simply put that out there early, since I truly believe that if there was no problem at all, as many would like to say, then AMD would not be working on this driver in the first place, since it'd be one huge waste of money. If you ignore that fact, that AMD is spending lots of cash on this issue, then yeah, I think you're mentally challenged, and I don't think there's anything wrong with saying so. I was directing my comments at any specific person, at all. I'm physically challenged, compared to someone who is stronger than I am. Why is someone being smarter, or being able to see things better than myself, a problem? Why is pointing it out a problem? It's simply science, in my books, form that perspective. I didn't say that's how it was, I said it could be construed...read all the text, not just the part that got you a bit upset.
 
Last edited:

MxPhenom 216

Corsair Fanboy
Joined
Aug 31, 2010
Messages
12,147 (4.50/day)
Likes
3,708
Location
Seattle, WA
System Name The Battlestation
Processor Intel Core i7 4770k @ 4.2GHZ 1.275v
Motherboard MSi Z97 Gaming 5
Cooling EK Supremacy w/ EK Coolstream PE360
Memory G. Skill Trident X 16Gb (4x4GB) 2400mhz @ 1.65v
Video Card(s) MSi GTX1070 Gaming X 8GB @ 2GHz
Storage Samsung 830 128GB SSD, Crucial MX200 500GB, Seagate Barracuda 2TB (2x 1TB Partitions)
Display(s) Qnix QX2710 27" 2560 x 1440 PLS @ 100hz
Case Phantek Enthoo Evolv ATX TG
Audio Device(s) MSi Gaming AudioBoost ALC1150 w/ Sennheiser Game Ones
Power Supply Seasonic Flagship Prime Platinum 850
Mouse Steelseries Rival 310 w/ QCK Mass mat
Keyboard Corsair K70 w/ MX Browns and Red Backlit
Software Windows 10 Pro 64-Bit
Benchmark Scores Firestrike: 15439
#25
So your retarded if you don't notice frame latency issues with AMD? (I'm sorry if I'm twisting your words but saying mental(ly) challenged people don't see a difference, what are you smoking Dave?) I could say the same for people who overly focus on the whole frame latency (non) issue whilst MANY people with dual AMD GPU setups for years have not seen any noticeable/real world issues (Fraps/Ffaps and faps aside)

Not meaning to be a dick and apologies if that's not how you meant to come across though you really should read what you post before hand as it could be construed in a way that's offensive, and IF you can't see that, maybe it's your problem bro.
I don't think he meant to say people are retarded or anything. I think he more meant some people who are not noticing the issue, their eyes, and mind are as tuned as some.

I know for myself, I notice a lot of small video issues, most of my friends won't notice. I'm just more in tune with it because I've been playing PC games for so long.

EDIT: Or maybe he did mean what he said lol.