[gamegpu] Dragon Age Inquisition performance

Page 11 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

SPBHM

Diamond Member
Sep 12, 2012
5,062
414
126
I'm playing DAI with the same VGA I used for DAII, and it looks a lot better, and DAII had some performance issues running in DX11 mode.
I'm comparing a game from 2011 with a new one, the difference is bigger than what I would expect based on other games, moving to the frostbite engine made a big difference, the lighting, scale and detail
 
Aug 11, 2008
10,451
642
126
Well, I ran DA2 on a 2.2 ghz core 2 duo and 9800 GT, and am sure I used at least medium, and I think it was high, and dont recall any performance problems.

With DAI, I have a 3ghz i5 and a HD7770 and can only run medium, and it still is not particularly smooth.
 

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
Dragon Age 2 basically had a bunch of DirectX 11 effects pasted a bit sloppily into a game engine designed for DirectX 9. The performance cost for the full DX11 effects on "Very High" was large, and the visual return on that performance wasn't all that great. DAI and its engine are designed for DX11 effects. DA2 still had performance hiccups even on my old 7870. DAI's performance seems more optimized for the level of detail it provides. I guess that would affect how well it scales down, though. So basically, DAI scales up better, while DA2 scaled down better.
 
Last edited:

DiogoDX

Senior member
Oct 11, 2012
757
336
136
Well, I ran DA2 on a 2.2 ghz core 2 duo and 9800 GT, and am sure I used at least medium, and I think it was high, and dont recall any performance problems.

With DAI, I have a 3ghz i5 and a HD7770 and can only run medium, and it still is not particularly smooth.
DAI on low looks way better than DA2 on very high.
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
It seems to me that some people are expecting every new game to blow their socks off on all big name releases. Unfortunately, that just doesn't happen very often. The last time something totally shocked me was playing Morrowind, because of how real the water looked compared to any game I had played before.

I also believe RPG's may have some technical requirements that hold back some IQ settings a little.

I will say that DA:I's faces are a huge step forward from what I've seen in any game before (I haven't bought another AAA game this year). It would have been nice if the hair had done the same.
 

Squeetard

Senior member
Nov 13, 2004
815
7
76
This game is my new tops in graphics. I love love love it. Both technically and the art direction. I can't stop taking screen shots of the beautiful environments and characters. Was at the red templar stronghold on Storm Coast last night and thought it was the best dungeon ever. Fantastic use of tesselation, everything has depth, gorgeous assets like the columns and junk. Water dripping down the walls.
2 - 7970's in xfire, ultra everything and it flies. 50-120fps.

RPG's for me are all about immersion, not 1500 deep skill tree's I can not get enough of wandering around in this game.
 

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
It seems to me that some people are expecting every new game to blow their socks off on all big name releases. Unfortunately, that just doesn't happen very often. The last time something totally shocked me was playing Morrowind, because of how real the water looked compared to any game I had played before.

I also believe RPG's may have some technical requirements that hold back some IQ settings a little.

I will say that DA:I's faces are a huge step forward from what I've seen in any game before (I haven't bought another AAA game this year). It would have been nice if the hair had done the same.

The detail of the faces are good, definitely a step up from DAO and DA2. Though they do kind of have a wax figure-ish look at times, I wonder if there's any sub-surface scattering going on to counter that?

On the subject of hair, there's more fine detail to it than past games, like with Cole's hair. There are some spots of noticeable hair physics, like Sera's hair, but it's minimal. The next step up I guess would be applying the TressFX technology. That needs to see wider use.
 

Head1985

Golden Member
Jul 8, 2014
1,867
699
136
No surprises really as that's been the case for 2+ years now.



- 980's lead over 970 is very consistent in the latest games, suggesting 970 is clearly TMU and/or shader performance bottlenecked (both of those are cut by 20% from the 980).
GTX980 leads only because reference GTX970 is bad card with only 145w TDP and because that reference GTX980 operate in much much higher clocks.

GTX980 is faster only by 15% vs same clock GTX970.
Look at this video both GTX970 and 980 on same speed and if you look at FPS you can see its only 15% difference
Dragon age 15%
https://www.youtube.com/watch?v=lsrBhnBOlQo
far cry4 15%
https://www.youtube.com/watch?v=qD7sSg0d6hg
Shadow of mordor+tomb raider+thief
https://www.youtube.com/watch?v=hIgcGir2-1s
In all those games GTX970 on same clock is only 15% slower than GTX980.

Problem is in reviews they use those crap super slow GTX970 and then you can see some really crazy differences between them.
 
Last edited:

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
It's pretty much always the case that harvested chips tend to be great overclockers when you push the voltage and TDP up. It's just a shaky thing to base purchases on because you're kind of playing the chip lottery. So it's not surprising that the 970 can hit close to the 980.
 

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
So, back when I did my DAI benchmarks, I never investigated memory usage. Well, I just did a rough approximation of memory usage with MSI Afterburner. My system was what I have in my sig, and the settings were everything maxed out except MSAA was off (post process AA was on max), and I tested with textures set to "Ultra" and textures set to "Fade-touched" (the maximum). I tested by walking around the crossroads in the Hinterlands and walking around Haven.

With textures set to Ultra, I got between 1600-1800 MB used of VRAM. With textures set to Fade-Touched I got between 1700-1900 MB used of VRAM. VRAM never seemed to touch 2000, though, and the game itself seems to recognize 2027 MB of VRAM via an in-game console display.

So that's the data, and here's my interpretation: Maybe my thoughts that I was running into a VRAM bottleneck aren't so true? At least, the game is not constantly chomping at the bit for more than 2 GB of VRAM. The highest texture setting doesn't actually add detail to textures, from what I've read; it just expands addressable cache size to help reduce pop-in. I guess you could push VRAM usage over the 2 GB threshold by both using fade-touched textures and turning on MSAA, which I didn't try because my 270X doesn't have the horsepower for it anyways. But when you actually consider the assets of the game and what it takes to render then, my findings indicate that at 1080P, Dragon Age Inquisition does not require more than 2 GB to get the most image detail out if it.
 

Deders

Platinum Member
Oct 14, 2012
2,401
1
91
I'm sure I've seen it use about 2.5GB on my 780, even before I was able to use AA.
 

Udgnim

Diamond Member
Apr 16, 2008
3,675
120
106
DA:I can load 3GB+ into VRAM on my R9 290 @ 1080

whether it is actually using 3GB+ is another question
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
So, back when I did my DAI benchmarks, I never investigated memory usage. Well, I just did a rough approximation of memory usage with MSI Afterburner. My system was what I have in my sig, and the settings were everything maxed out except MSAA was off (post process AA was on max), and I tested with textures set to "Ultra" and textures set to "Fade-touched" (the maximum). I tested by walking around the crossroads in the Hinterlands and walking around Haven.

With textures set to Ultra, I got between 1600-1800 MB used of VRAM. With textures set to Fade-Touched I got between 1700-1900 MB used of VRAM. VRAM never seemed to touch 2000, though, and the game itself seems to recognize 2027 MB of VRAM via an in-game console display.

So that's the data, and here's my interpretation: Maybe my thoughts that I was running into a VRAM bottleneck aren't so true? At least, the game is not constantly chomping at the bit for more than 2 GB of VRAM. The highest texture setting doesn't actually add detail to textures, from what I've read; it just expands addressable cache size to help reduce pop-in. I guess you could push VRAM usage over the 2 GB threshold by both using fade-touched textures and turning on MSAA, which I didn't try because my 270X doesn't have the horsepower for it anyways. But when you actually consider the assets of the game and what it takes to render then, my findings indicate that at 1080P, Dragon Age Inquisition does not require more than 2 GB to get the most image detail out if it.

What were the rest of your settings? I personally found tessellation behaved as if it pushed my VRAM usage more than textures did. Or at least, it sent me over the top, particularly in places like Crossroads and Redcliff.
 

Pneumothorax

Golden Member
Nov 4, 2002
1,181
23
81
... But when you actually consider the assets of the game and what it takes to render then, my findings indicate that at 1080P, Dragon Age Inquisition does not require more than 2 GB to get the most image detail out if it.

You're right, my son's rig that's hooked up to a 1080P TV never hits over 3gb on DAI. Mine, that's running at 1600P in CF, will go over 3GB of usage at times.
 

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
What were the rest of your settings? I personally found tessellation behaved as if it pushed my VRAM usage more than textures did. Or at least, it sent me over the top, particularly in places like Crossroads and Redcliff.

As I said, all my settings were maxed out, including tessellation, except MSAA was off and I tested between textures on "ultra" and textures on "fade touched".

I suppose it's possible that with textures on "fade-touched", it could be automatically switching out somewhat to system RAM before the VRAM ever hits over 2000 MB. But on ultra textures, VRAM usage was distinctly sitting below 1900 MB, usually under 1800 MB.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
So, back when I did my DAI benchmarks, I never investigated memory usage. Well, I just did a rough approximation of memory usage with MSI Afterburner. My system was what I have in my sig, and the settings were everything maxed out except MSAA was off (post process AA was on max), and I tested with textures set to "Ultra" and textures set to "Fade-touched" (the maximum). I tested by walking around the crossroads in the Hinterlands and walking around Haven.

With textures set to Ultra, I got between 1600-1800 MB used of VRAM. With textures set to Fade-Touched I got between 1700-1900 MB used of VRAM. VRAM never seemed to touch 2000, though, and the game itself seems to recognize 2027 MB of VRAM via an in-game console display.

So that's the data, and here's my interpretation: Maybe my thoughts that I was running into a VRAM bottleneck aren't so true? At least, the game is not constantly chomping at the bit for more than 2 GB of VRAM. The highest texture setting doesn't actually add detail to textures, from what I've read; it just expands addressable cache size to help reduce pop-in. I guess you could push VRAM usage over the 2 GB threshold by both using fade-touched textures and turning on MSAA, which I didn't try because my 270X doesn't have the horsepower for it anyways. But when you actually consider the assets of the game and what it takes to render then, my findings indicate that at 1080P, Dragon Age Inquisition does not require more than 2 GB to get the most image detail out if it.

When I had my 2GB 670s the game stuttered with fade touched textures. The same settings and everything with the next step down and the stuttering went away. The benchmark didn't show any FPS change but it was a noticeable hitching in certain areas and the fps counter had a drop for sure. Now with my 970s that went away completely. I wouldn't know if it's the VRAM or not but it could be, but the reported usage could also be cached and not actually in use.
 

Deders

Platinum Member
Oct 14, 2012
2,401
1
91
I didn't get stuttering at all with a 3GB 780 but the stuttering is one thing in the driver notes that Nvidia still hasn't dealt with yet.

There is always the possibility that the more memory you have the more can be cached just in case it's needed.
 

Red Hawk

Diamond Member
Jan 1, 2011
3,266
169
106
With the latest Inquisition patch and AMD drivers, I decided to do a little testing on my brother's PC (stock Q6600, Radeon HD 5770 1 GB @ 890/1300, 4 GB 667 MHz RAM, Windows 7). I had MSI Afterburner running in the background to monitor GPU and VRAM usage, which I hadn't done before with his system.

The game really does seem CPU limited, using all 4 cores. The monitor indicated they all stayed above at least 80% usage while running the game. Walking around the Hinterlands, the game fluctuated around 40 frames per second -- when there were no NPCs. In a high-NPC area like the Hinterlands Crossroads, the framerate plummeted below 20 FPS. Checking the monitor, GPU usage also went down to around 70% or 60% in the Crossroads, while it stayed above 90% outside the Crossroads. Low GPU usage is an indicator that the CPU isn't feeding the GPU enough frames to fully use it.

So, Inquisition should see benefit from Mantle in CPU performance with something like a Q6600. We haven't seen much GPU benefit from Mantle with Inquisition, but CPU is supposed to be the bigger benefit with Mantle. Of course, the 5770 doesn't support Mantle, but my brother might be getting a Radeon 260X in the near future (along with 4 sticks of 2 GB 800 MHz DDR2 RAM), so I might be able to test it. We'll see.
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
A month or 2 ago, when I was playing, I noticed some interesting behavior in those areas.

I was experimenting with the " -GameTime.MaxSimFps 60 -GameTime.ForceSimRate 60+" command options. I turned it up to 120 FPS. The game was running smooth, until I hit the Redcliff. All or a sudden, the game was running in slow motion, even though I had 45-60ish FPS. I dropped the ForceSimRate to 60, and the game ran at normal speed.

I'm guessing that the SimRate is how fast it runs physics, AI and other simulation stuff, and it will slow down everything to make sure it completes its tasks. When I set it to 120, my CPU could not keep up, and just ran things in slow motion, as it had to complete all those tasks and my CPU was just too slow for it.
 

alcoholbob

Diamond Member
May 24, 2005
6,349
423
126
I think open world games are going to get a huge boost in performance when DX12 comes around. In a lot of areas in Dragon Age in Inquisition (large outdoor areas) I suspect are CPU limited.

Going from a GTX 980 to a Titan X at 1400MHz (about a 60% increase in processing power) I saw about a 50% performance increase in cutscenes and small maps. On big maps it's only about 20% faster, and overclocking or downclocking the card to stock (50% slower) barely changed my framerate, maybe by 1 or 2 fps.
 

escrow4

Diamond Member
Feb 4, 2013
3,339
122
106
With the latest Inquisition patch and AMD drivers, I decided to do a little testing on my brother's PC (stock Q6600, Radeon HD 5770 1 GB @ 890/1300, 4 GB 667 MHz RAM, Windows 7). I had MSI Afterburner running in the background to monitor GPU and VRAM usage, which I hadn't done before with his system.

The game really does seem CPU limited, using all 4 cores. The monitor indicated they all stayed above at least 80% usage while running the game. Walking around the Hinterlands, the game fluctuated around 40 frames per second -- when there were no NPCs. In a high-NPC area like the Hinterlands Crossroads, the framerate plummeted below 20 FPS. Checking the monitor, GPU usage also went down to around 70% or 60% in the Crossroads, while it stayed above 90% outside the Crossroads. Low GPU usage is an indicator that the CPU isn't feeding the GPU enough frames to fully use it.

So, Inquisition should see benefit from Mantle in CPU performance with something like a Q6600. We haven't seen much GPU benefit from Mantle with Inquisition, but CPU is supposed to be the bigger benefit with Mantle. Of course, the 5770 doesn't support Mantle, but my brother might be getting a Radeon 260X in the near future (along with 4 sticks of 2 GB 800 MHz DDR2 RAM), so I might be able to test it. We'll see.

My 5930K @ 3.7GHz pushed my 780Ti @ 1215MHz to 99% as far as I could see throughout the game, though I did notice dips in Redcliffe and Crossroads (40-50FPS, as well as random drops to sub 40s in that forested area). RPGs are frequently CPU heavy anyway with all those calculations/AI in the background anyway. This was before all these patches though. Stupid Bioware, patch it properly so Day 1's can play it properly . . . . in November 2014.
 

TheELF

Diamond Member
Dec 22, 2012
4,027
753
126
Hey,can anybody with a monster rig,since I see you having those,check out if this still gives you any benefit at all??
Would be very interesting.
Basically you have to lower the games priority to background or even idle and you get a nice boost,at least for me that was the case.
https://www.youtube.com/watch?v=Tlm2guF4nxk
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |