My 3090 Settings for my G2 (2.7.11)

[Stefan Gebhardt]

Many thanks to speed-of-heat for his guideance and work! I am new to DCS (even if I owned it for years) and thought I will restart again when I got a new computer. I did some testing and thought it make sense to share my results. Hardware setup is listed in the signature. 

Benchmark track: Performance Track 276

Benchmark conditions:

• Recording via fpsVR, started right away and ended when the plane reached a hight of 500ft over ground (Frametimes SteamVR from 120.9315 to 416.8695.
• No change to the joystick, thrust or vr view during the flight
• DCS 2.7.6.16436
• Overall setup following speed-of-heat recommendations

DCS settings:

WRP settings:

SteamVR settings:

Benchmark run 1 - No mods

Benchmark run 2 - DCS VR Shaders mod 2.7.7.14727 + OpenVR FSR V1.3

Edited November 4, 2021 by Stefan Gebhardt

[X_legio]

Very nice suggestions.

Can someone can explain me why I have your same Steam and DCS settings but a totally different resolution ? 
Also in DCS the PID is 1.0

Edited November 5, 2021 by X_legio

[dureiken]

On 10/27/2021 at 4:30 PM, speed-of-heat said:

2) one of the main differences between G1 and G2 is that the G2 operates at 150% of it physical resolution to accommodate barrel distortion, which gives the setting 3168 x 3096 at 100% in Steam's Resolution per Eye for the G2

 

[speed-of-heat]

I have seen differences in the past, never been able to track down why.

[daddio007]

On 11/2/2021 at 3:51 AM, ShaunX said:

That's not quite my understanding. An FSR renderscale of .8 means that it will reduce the SteamVR CRM setting by a factor of .8, and then use its own algorythms to upscale back to that SteamVR resolution. (Though only over a central area of the display as defined in its radius parameter).

The effect of this is to lower your GPU and CPU framerates while still achieving the same CRM as set in SteamVR (albeit with perhaps some increased shimmering). You specifically don't need to raise SteamVR CRM in conjunction to achieve this.

What people have found however, is that the significantly lower GPU and CPU framerates now give them the bandwidth to be able to run at a raised 100% SteamVR and thus achieve even further clarity.

Personally, I leave SteamVR CRM at 80% as I would rather use those lower framerates to allow me to run Reshade sharpening (and with FSR "sharpness": 0.0) as I find that gives me better clarity (certainly in terrain detail) than 100% CRM and it also allows me to maintain 45FPS low over Guam !

 

 

Hey guys!

I have had pretty good success following this guide but there are places in some of the maps that cause some serious interrupts. One of these is over Guam so this post caught my eye. This is what I get over certain parts of Guam:

It's enough that sometimes Steam VR will crash with this in the Steam vr logs: Closing pipe VR_ServerPipe_5704 because it was broken from the other end

Any ideas troubleshooting this? I've seem everything from rolling back drivers to disable motion repro to re-install Steam.

Mostly it is fine and there are places where I'm running at 90fps with under 10ms frametimes on GPU and CPU. But certain places where there might be a lot of certain objects it nose dives in a hurry!

Core I9-9900KF, 32 gb ram, NVM drives, RTX 3090, almost same specs as SOH except still W10.

Thanks in advance for any suggestions!

[speed-of-heat]

Alas no, some maps do give worse perf than others, sorry to say. And the Marianas is a nasty offender anytime you are overland.

[Stefan Gebhardt]

10 hours ago, daddio007 said:

 

Hey guys!

I have had pretty good success following this guide but there are places in some of the maps that cause some serious interrupts. One of these is over Guam so this post caught my eye. This is what I get over certain parts of Guam:

It's enough that sometimes Steam VR will crash with this in the Steam vr logs: Closing pipe VR_ServerPipe_5704 because it was broken from the other end

Any ideas troubleshooting this? I've seem everything from rolling back drivers to disable motion repro to re-install Steam.

Mostly it is fine and there are places where I'm running at 90fps with under 10ms frametimes on GPU and CPU. But certain places where there might be a lot of certain objects it nose dives in a hurry!

Core I9-9900KF, 32 gb ram, NVM drives, RTX 3090, almost same specs as SOH except still W10.

Thanks in advance for any suggestions!

The info for the usage of video memory in fpsVR puzzles me. 23 out of  24GB? I need to check this detail for my setup next time…

 

Edited November 7, 2021 by Stefan Gebhardt

[speed-of-heat]

yep pretty common on the newer maps; remember its showing allocated rather than used.. .but still 

[Stefan Gebhardt]

2 hours ago, speed-of-heat said:

yep pretty common on the newer maps; remember its showing allocated rather than used.. .but still 

Will check to confirm

[Burt]

5 hours ago, Stefan Gebhardt said:

Will check to confirm

Wow ! Same here.

[Taterbootz]

20 hours ago, daddio007 said:

 

Hey guys!

I have had pretty good success following this guide but there are places in some of the maps that cause some serious interrupts. One of these is over Guam so this post caught my eye. This is what I get over certain parts of Guam:

It's enough that sometimes Steam VR will crash with this in the Steam vr logs: Closing pipe VR_ServerPipe_5704 because it was broken from the other end

Any ideas troubleshooting this? I've seem everything from rolling back drivers to disable motion repro to re-install Steam.

Mostly it is fine and there are places where I'm running at 90fps with under 10ms frametimes on GPU and CPU. But certain places where there might be a lot of certain objects it nose dives in a hurry!

Core I9-9900KF, 32 gb ram, NVM drives, RTX 3090, almost same specs as SOH except still W10.

Thanks in advance for any suggestions!

I've been running into the same issue with SteamVR lately.  When I chased it down in the SteamVR logs it was an audio issue when I was also heavily using the GPU, so I disabled "Loudness Equalization" from the headset properties and that seems to have fixed it for me.  Check all of the steamVR logs to see what was causing it then you might be able to track it down better.

[speed-of-heat]

 

32 minutes ago, Taterbootz said:

I've been running into the same issue with SteamVR lately.  When I chased it down in the SteamVR logs it was an audio issue when I was also heavily using the GPU, so I disabled "Loudness Equalization" from the headset properties and that seems to have fixed it for me.  Check all of the steamVR logs to see what was causing it then you might be able to track it down better.

I can't find a Loudness equalisation setting... what headset are you using?

and where did you find it?

[speed-of-heat]

nvm .. found it windows -> control panel -> sound -> manage audio devices ... mine was set off anyway ,so i dont think it was that causing increased gpu memory usage 

[Taterbootz]

Yeah I cant see it causing increased GPU mem usage at all, but it could just be the straw that breaks the fragile WMR/SteamVR relationships back for some reason.  I've been struggling with issues with steamVR crashing a lot myself. 

Actually just crashed when I started up my controller, so I found that I had two profiles for controllers in the settings and my drivers werent up to date.  Currently its working good.

 

[speed-of-heat]

Alderlake,

I'm still on DDR4 and most of the reviews seem to say that in some edge cases you will get some perf improvement from DDR5, to date none of them are gaming scenarios.

Overall I got a huge performance improvement in terms of CPU frametime (20-25%) , so if you are bottle necked at the CPU then it is a great upgrade.  I did some additional HAGS testing , and the behaviour is very odd... it seems to lock at 60hz and I wonder if some of the people saying they are locked at 60 in game and 90 in menus have hags enabled because that is exactly the behaviour I  am seeing:

Frametime in ms		Delta From Baseline
Ver	Test	Baseline	GPU	CPU	GPU	CPU	Changes
2.7.7.15038	100	Test 97	10.61	8.81			Shader Cache to 10Gb
2.7.7.15038	101	Test 100	10.66	6.92	-0.05	1.89	Alderlake
2.7.7.15038	102	Test 100	10.61	6.78	0.00	2.03	HAGS
2.7.7.15038	103	Test 100	10.68	7.01	-0.07	1.81	HAGS

even though the frame times are identical it seems to be quite a stable 60hz, but is very choppy when viewed at 90hz, i have not tried running my headset at 60 and seeing what happens, the increase in GPU frametime are almost certainly runtime variance. 

In terms of overclock, i'm running the AI OC from Asus which gives me a clock speed of 5.1, with a 5.3 turbo on the p-cores and 3.8 on the e-cores

So was it worth it, I'd say yes for me I skipped over 10 and 11 , and I seem to be running cooler now as well 

Note I had 3dmigoto running (i forgot so sue me) for this minus that  i now get another 0.7ms on cpu frame time... 

 

			Frametime in ms		Delta From Baseline
Ver	Test	Baseline	GPU	CPU	GPU	CPU	Changes
2.7.7.15038	100	Test 97	10.61	8.81	-0.55	-0.73	Shader Cache to 10Gb
2.7.7.15038	101	Test 100	10.66	6.92	-0.05	1.89	Alderlake
2.7.7.15038	102	Test 100	10.61	6.78	0.00	2.03	HAGS
2.7.7.15038	103	Test 100	10.68	7.01	-0.07	1.81	HAGS
2.7.7.15038	104	Test 100	10.72	6.25	-0.11	2.56	minus 3dmigimoto

Edited November 9, 2021 by speed-of-heat
-3dmigimoto

[ShaunX]

3 minutes ago, speed-of-heat said:

Alderlake,

I'm still on DDR4 and most of the reviews seem to say that in some edge cases you will get some perf improvement from DDR5, to date none of them are gaming scenarios.

Overall I got a huge performance improvement in terms of CPU frametime (20-25%) , so if you are bottle necked at the CPU then it is a great upgrade.  I did some additional HAGS testing , and the behaviour is very odd... it seems to lock at 60hz and I wonder if some of the people saying they are locked at 60 in game and 90 in menus have hags enabled because that is exactly the behaviour I  am seeing:

Frametime in ms		Delta From Baseline
Ver	Test	Baseline	GPU	CPU	GPU	CPU	Changes
2.7.7.15038	100	Test 97	10.61	8.81			Shader Cache to 10Gb
2.7.7.15038	101	Test 100	10.66	6.92	-0.05	1.89	Alderlake
2.7.7.15038	102	Test 100	10.61	6.78	0.00	2.03	HAGS
2.7.7.15038	103	Test 100	10.68	7.01	-0.07	1.81	HAGS

even though the frame times are identical it seems to be quite a stable 60hz, but is very choppy when viewed at 90hz, i have not tried running my headset at 60 and seeing what happens, the increase in GPU frametime are almost certainly runtime variance. 

In terms of overclock, i'm running the AI OC from Asus which gives me a clock speed of 5.1, with a 5.3 turbo on the p-cores and 3.6 on the e-cores

So was it worth it, I'd say yes for me I skipped over 10 and 11 , and I seem to be running cooler now as well 

Very nice bump indeed. I feel a bout of CPU envy coming on

 

[speed-of-heat]

6 minutes ago, ShaunX said:

Very nice bump indeed. I feel a bout of CPU envy coming on

 

It is good, and I am not ungrateful for it, but, functionally i'm now GPU bound... on a 3090 (let that sink in for a moment) ... 

[Hiob]

4 minutes ago, speed-of-heat said:

It is good, and I am not ungrateful for it, but, functionally i'm now GPU bound... on a 3090 (let that sink in for a moment) ... 

Imagine the opposite, being cpu bound on the latest and greatest cpu-genaration... would be even more insane.

Btw. if I interpret your numbers correctly, you weren't cpu bound with your last cpu already, or were you?

[ShaunX]

7 minutes ago, speed-of-heat said:

It is good, and I am not ungrateful for it, but, functionally i'm now GPU bound... on a 3090 (let that sink in for a moment) ... 

That's pretty scary!

However, no need to worry, because according to ED, 2.7 and now 2.7.7 releases didn't place any more demand on CPU or GPU LOL!

Start saving those pennies again.

[speed-of-heat]

1 minute ago, ShaunX said:

That's pretty scary!

However, no need to worry, because according to ED, 2.7 and now 2.7.7 releases didn't place any more demand on CPU or GPU LOL!

Start saving those pennies again.

lol but for what !!!

[ShaunX]

2 minutes ago, speed-of-heat said:

lol but for what !!!

erm, at this rate, psychiatric counselling fees perhaps

[speed-of-heat]

[RvL]

1 hour ago, speed-of-heat said:

Alderlake,

I'm still on DDR4 and most of the reviews seem to say that in some edge cases you will get some perf improvement from DDR5, to date none of them are gaming scenarios.

Overall I got a huge performance improvement in terms of CPU frametime (20-25%) , so if you are bottle necked at the CPU then it is a great upgrade.  I did some additional HAGS testing , and the behaviour is very odd... it seems to lock at 60hz and I wonder if some of the people saying they are locked at 60 in game and 90 in menus have hags enabled because that is exactly the behaviour I  am seeing:

Frametime in ms		Delta From Baseline
Ver	Test	Baseline	GPU	CPU	GPU	CPU	Changes
2.7.7.15038	100	Test 97	10.61	8.81			Shader Cache to 10Gb
2.7.7.15038	101	Test 100	10.66	6.92	-0.05	1.89	Alderlake
2.7.7.15038	102	Test 100	10.61	6.78	0.00	2.03	HAGS
2.7.7.15038	103	Test 100	10.68	7.01	-0.07	1.81	HAGS

even though the frame times are identical it seems to be quite a stable 60hz, but is very choppy when viewed at 90hz, i have not tried running my headset at 60 and seeing what happens, the increase in GPU frametime are almost certainly runtime variance. 

In terms of overclock, i'm running the AI OC from Asus which gives me a clock speed of 5.1, with a 5.3 turbo on the p-cores and 3.8 on the e-cores

So was it worth it, I'd say yes for me I skipped over 10 and 11 , and I seem to be running cooler now as well 

 

Interesting. That sounds just like my issue while HAGS is disabled. But gonna switch it again a few times in case it didnt liked it being switched just once.

[speed-of-heat]

OP updated with Alderlake

[Harlikwin]

1 hour ago, speed-of-heat said:

OP updated with Alderlake

Can you post details of your alderlake setup? Hardware and win10 or 11?

I'm still trying to find RAM to build mine, but 30% drop in CPU frametime seems very promising.