Jump to content
Sign in to follow this  
iFlySimX

Limited by MainThread

Recommended Posts

4 hours ago, RJC68 said:

I am curious how you got your 12900K running at 100% in MSFS or am I misunderstanding you. I just rebuilt my PC with an i7-12700K & EVGA Ultra 3080TI FW3 and even with no fps lock in place I rarely see the CPU usage above 30%. With the Fenix A320 I still lock the FPS to 30 and with Ultra settings I am currently seeing around 15% CPU@56 degrees and 53% GPU@66 degrees. I also undervolted the 3080TI which makes a huge difference in temps and fan speeds with little to no drop in performance,

I am hoping SU10 will bring some of the performance back, I can hit 60fps in the Fenix but need to dial back some of the settings and even then it isn't consistent especially at busy airports and the GPU temps climb pretty quickly. Last night I flew the Fenix into London Heathrow with live traffic and Ultra settings @ 30FPS, TLOD 200 & OLOD 200 and it was butter smooth, not a single stutter or slowdown right too the gate. It would be really handy if we were able to create several settings presets for different aircraft/situations and easily switch between them.

What CPU did you upgrade from?  For me the upgrade to 12th gen (as well as a fresh install of windows11) has been awesome.

  • Like 1

i9 9900K at 4.9ghz, MSI RTX 3080, 32 G RAM, (3) 1TB SSD 

Share this post


Link to post
Share on other sites
1 minute ago, Poppingcork said:

What CPU did you upgrade from?  For me the upgrade to 12th gen (as well as a fresh install of windows11) has been awesome.

I had an i5-9600K & 1080TI, this served me well in X-Plane 11 and then MSFS 2020 but my new PC is next level.

The funny thing is when I went to buy the CPU & Motherboard the guy at the store talked me out of the i7 and I came away with the i5-12600K. I slept on it overnight and decided I wanted the i7-12700K as I plan on also doing some video editing so I went back and picked it up. From what I am seeing the 12600K probably would have sufficed. I have been thrilled so far with my new build, the 3080TI is a beast as well. As I mentioned earlier in the Fenix I chose to stick to 30fps but everything is maxed out including road traffic, ships and airport traffic and the sim looks spectacular on my 4K TV. I haven't smiled this much in a long time lol

  • Like 3
  • Upvote 1

 

Richard

i7-12700K | Noctua NH-D15S Black Version | MSI Pro Z690 - A | 32 GB DDR4 3600 | Gigabyte Gaming OC 4090 | 1TB WD Blue NMVe (MSFS 2020) | 500 GB WD Black Gen 4 NVMe | 4TB WD Black Conventional | Fractal Design Torrent Case | Seasonic 1000W Gold Plus PSU | Thrustmaster Boeing Yoke | Honeycomb Throttle | Airbus Side Stick | Virpil Rudder Pedals | Sony X90K 55 Inch TV |

mmBbmS1.png

 

Share this post


Link to post
Share on other sites
11 hours ago, molleh said:

my performance findings with a 9900k, 3080 ti and a 4k 120hz gsync/freesync display (no overclocks on anything):

- hyperthreading off increases overall smoothness (more consistent frame pacing, less stuttering.)

- using nvidia control panel to set the msfs profile to 'fixed refresh' instead of gsync and selecting '20 fps' for the fps limit in msfs' graphics settings (i.e. 1/3rd vsync and 40 fps cap) results increases overall smoothness (more consistent frame pacing, less stuttering.) this doesn't make sense since selecting either '30 fps' or '20 fps' should disable gsync anyway since gsync isn't supported while using 1/2, 1/3, 1/4 refresh rate vsync settings, but i've done back to back testing repeatedly and it absolutely does make a fairly significant difference; settings need to be adjusted to maintain 40 fps though, as dropping below that causes the normal vsync uneven frame pacing. i've also been using 'prefer max performance' for the power setting, although i do need to go back and test it with 'normal' to see if it could save a bit on power consumption.

- i also use nvcp to set 'threaded optimization' and 'triple buffering' to off; contrary to popular belief, 'threaded optimization' isn't a catch-all term, but specifically refers to handling of vsync (not worth typing out here, but you can research it for a better understanding.)

- terrain lod and ai traffic are the biggest factors in main thread performance, and to a lesser extent, airport traffic (service/fuel trucks.) terrain lod of 100 is the minimum i find acceptable in terms of graphics, because the autogen draw distance and ortho lod falls too far below 100. i can often run terrain lod at 200, but it does result in much more frequent drops below 40 fps at bigger airports, especially with the fenix a320. even terrain lod 100 is not even to maintain a 100% lock to 40 fps at all times at a few problematic airports like KDFW and KORD.

1) No measurements, no numbers, just a subjective statement. As usual. 

2) Framerate capped at 20? No thanks. No wonder you see less stuttering, at 20 fps it's a continuous stuttering. With GSync then, it doesn't make sense to cap the framerate to a submultiple of the screen refresh frequency because the refresh will always be equal to the framerate. Unless you have a GSync "compatible" monitor, in that case VRR stops working below 60 or 48 Hz.

Prefer max performance setting is another placebo: your CPU and GPU work at the highest possible frequency as long as there is a significant load (you can verify it with Riva Tuner overlay). I use Balanced and Normal for the CPU and GPU respectively and MSFS is smooth as silk.

3) Other placebo.

4) I personally find TLOD 200 the best compromise. I never drop below 32-33 fps on the ground with airliners at heavy airports and 48 fps on the air (capped in order to keep the 3090 quiet) - which with GSync is very smooth.  

16 hours ago, Seth2021 said:

As i mentioned i am not a fan of  HT off etc. I made some test flights last night and indeed DX12 and HT off spreads more "equally" the work to threads. With HT enabled there is always the last core hammered with all limitations.

Check attached process lasso preview with rules i described on my previous post. 

also i really aware the placebo effect, and its not.

spacer.png

In this screen HT is NOT disabled. You have 16 active logical cores.

To disable HT you have to enter the BIOS.

11 hours ago, iFlySimX said:

To you think there is point of me OC my i9-9900k to 4.8ghz?

 

I don't get the question, according to your signature it's already at 5.0 GHz.

The performance gain is modest, less than proportional to the frequency increase... so in the order of 2 fps.

  • Like 2
  • Upvote 2

7800X3D | 32 GB DDR5-6000 | RTX 3090 | Acer Predator X34P GSync | Tobii Eye Tracker 5 | Completed all achievements 😛 https://i.postimg.cc/DyjR8mzG/image.png

Share this post


Link to post
Share on other sites
1 hour ago, MrFuzzy said:

In this screen HT is NOT disabled. You have 16 active logical cores.

To disable HT you have to enter the BIOS.

i know HOW  to disable HT from Bios.

https://bitsum.com/helpie_faq/should-i-disable-smt-hyper-threading/#:~:text=Process Lasso has the capability,are available to target processes.

I disabled HT for MSFS not for my whole system.

 

 

Edited by Seth2021
  • Upvote 1

Intel Core i9-13900K | ASUS ROG STRIX GeForce RTX 4090 GAMING OC 24GB | ASUS ROG MAXIMUS Z790 HERO | DDR5 64GB 6000-30 Trident Z5 RGB | Corsair ICUE H170i Elite Capellix RGB | Corsair 7000D Airflow ASUS ROG Thor Platinum II 1200 Watt | Samsung SSD 990 Pro NVMe M.2 2TB & 1TB | Alienware AW3423DW | Asus ROG Swift PG279Q 27" Gaming Monitor | VKB-Sim Gladiator Mk.II | Thrustmaster TCA Quadrant Airbus Edition

Share this post


Link to post
Share on other sites

to be clear, i am against tweaks, placebo effects, HT off or on etc. But my mission right now on SU10 and especially on DX12 which is still in beta is to test.

with  Process Lasso method of  HT ON  the results on cpu core/threads looks like that:

bCjAVfT.jpg

 

with Process Lasso method of HT OFF , or Hyper Placebo Threading OFF , core usage is like that:

21wQila.jpg

 


Intel Core i9-13900K | ASUS ROG STRIX GeForce RTX 4090 GAMING OC 24GB | ASUS ROG MAXIMUS Z790 HERO | DDR5 64GB 6000-30 Trident Z5 RGB | Corsair ICUE H170i Elite Capellix RGB | Corsair 7000D Airflow ASUS ROG Thor Platinum II 1200 Watt | Samsung SSD 990 Pro NVMe M.2 2TB & 1TB | Alienware AW3423DW | Asus ROG Swift PG279Q 27" Gaming Monitor | VKB-Sim Gladiator Mk.II | Thrustmaster TCA Quadrant Airbus Edition

Share this post


Link to post
Share on other sites
3 minutes ago, Seth2021 said:

to be clear, i am against tweaks, placebo effects, HT off or on etc. But my mission right now on SU10 and especially on DX12 which is still in beta is to test.

with  Process Lasso method of  HT ON  the results on cpu core/threads looks like that:

bCjAVfT.jpg

 

with Process Lasso method of HT OFF , or Hyper Placebo Threading OFF , core usage is like that:

21wQila.jpg

 

Not sure why with HT on it looks like that for you. I have HT on and mine looks like your bottom pic.

Edited by Ianrivaldosmith

Share this post


Link to post
Share on other sites
8 minutes ago, Ianrivaldosmith said:

Not sure why with HT on it looks like that for you. I have HT on and mine looks like your bottom pic.

Thats why i want to check and test to understand how MSFS works with my cpu etc.

this is the famous hammered Main thread situation, based on MSFS official forum majority of users have the same result.

 


Intel Core i9-13900K | ASUS ROG STRIX GeForce RTX 4090 GAMING OC 24GB | ASUS ROG MAXIMUS Z790 HERO | DDR5 64GB 6000-30 Trident Z5 RGB | Corsair ICUE H170i Elite Capellix RGB | Corsair 7000D Airflow ASUS ROG Thor Platinum II 1200 Watt | Samsung SSD 990 Pro NVMe M.2 2TB & 1TB | Alienware AW3423DW | Asus ROG Swift PG279Q 27" Gaming Monitor | VKB-Sim Gladiator Mk.II | Thrustmaster TCA Quadrant Airbus Edition

Share this post


Link to post
Share on other sites
1 hour ago, Seth2021 said:

Thats why i want to check and test to understand how MSFS works with my cpu etc.

this is the famous hammered Main thread situation, based on MSFS official forum majority of users have the same result.

 

I should add, in DX 11 my cpu looks like your top pic. It's only in Dx12 that it spreads evenly. Both with HT on. 

Share this post


Link to post
Share on other sites
3 hours ago, Ianrivaldosmith said:

I should add, in DX 11 my cpu looks like your top pic. It's only in Dx12 that it spreads evenly. Both with HT on. 

Yes, that was also my first impression in the beginning, after watching Process Lasso cpu monitor usage while using MSFS, the Main Thread hammering was back even on DX12. Will need to check and test more

 


Intel Core i9-13900K | ASUS ROG STRIX GeForce RTX 4090 GAMING OC 24GB | ASUS ROG MAXIMUS Z790 HERO | DDR5 64GB 6000-30 Trident Z5 RGB | Corsair ICUE H170i Elite Capellix RGB | Corsair 7000D Airflow ASUS ROG Thor Platinum II 1200 Watt | Samsung SSD 990 Pro NVMe M.2 2TB & 1TB | Alienware AW3423DW | Asus ROG Swift PG279Q 27" Gaming Monitor | VKB-Sim Gladiator Mk.II | Thrustmaster TCA Quadrant Airbus Edition

Share this post


Link to post
Share on other sites
13 hours ago, MrFuzzy said:

1) No measurements, no numbers, just a subjective statement. As usual. 

2) Framerate capped at 20? No thanks. No wonder you see less stuttering, at 20 fps it's a continuous stuttering. With GSync then, it doesn't make sense to cap the framerate to a submultiple of the screen refresh frequency because the refresh will always be equal to the framerate. Unless you have a GSync "compatible" monitor, in that case VRR stops working below 60 or 48 Hz.

Prefer max performance setting is another placebo: your CPU and GPU work at the highest possible frequency as long as there is a significant load (you can verify it with Riva Tuner overlay). I use Balanced and Normal for the CPU and GPU respectively and MSFS is smooth as silk.

3) Other placebo.

4) I personally find TLOD 200 the best compromise. I never drop below 32-33 fps on the ground with airliners at heavy airports and 48 fps on the air (capped in order to keep the 3090 quiet) - which with GSync is very smooth.  

In this screen HT is NOT disabled. You have 16 active logical cores.

To disable HT you have to enter the BIOS.

I don't get the question, according to your signature it's already at 5.0 GHz.

The performance gain is modest, less than proportional to the frequency increase... so in the order of 2 fps.

well, at least you demonstrated that you don't understand how the "frame rate limit" setting in msfs works.

Share this post


Link to post
Share on other sites
10 hours ago, molleh said:

well, at least you demonstrated that you don't understand how the "frame rate limit" setting in msfs works.

You demonstrated that you cannot quote a post. Explain please.


7800X3D | 32 GB DDR5-6000 | RTX 3090 | Acer Predator X34P GSync | Tobii Eye Tracker 5 | Completed all achievements 😛 https://i.postimg.cc/DyjR8mzG/image.png

Share this post


Link to post
Share on other sites

P.S. If you refer to the MSFS graphics menu, it has been corrected and it's not "20-30-60" regardless of the screen refresh anymore.


7800X3D | 32 GB DDR5-6000 | RTX 3090 | Acer Predator X34P GSync | Tobii Eye Tracker 5 | Completed all achievements 😛 https://i.postimg.cc/DyjR8mzG/image.png

Share this post


Link to post
Share on other sites
4 hours ago, MrFuzzy said:

You demonstrated that you cannot quote a post. Explain please.

Don't get drawn under the bridge Fuzzy!  You know that is where they live!  :biggrin:

  • Like 3

Call me Bob or Rob, I don't mind, but I prefer Rob.

I like to trick airline passengers into thinking I have my own swimming pool in my back yard by painting a large blue rectangle on my patio.

Intel 14900K in a Z790 motherboard with water cooling, RTX 4080, 32 GB 6000 CL30 DDR5 RAM, W11 and MSFS on Samsung 980 Pro NVME SSD's.  Core Isolation Off, Game Mode Off.

Share this post


Link to post
Share on other sites
On 7/11/2022 at 9:44 PM, micstatic said:

Exactly why I went with the 12700. Easier to keep cool also. 

This is something a lot of people overlook, especially 3080s. Mine was thermal throttling a lot, and in some games crashing out completely because it was getting too hot. 
 

I finally broke down and decided to do a water loop for my CPU/GPU and it is the best investment I’ve made in a long time


Tyson Rose

Share this post


Link to post
Share on other sites

I was going crazy trying to diagnose a cpu limited condition.  The most noticeable symptoms were severe lag spikes when view panning and severe lag when zooming the world map.   I'd see 1% lows in the single digits with the max fps around 60.

I kept thinking it was a problem with either the chipset drivers, the CPU power plan,  or GPU drivers.   I tried GPU and chipset driver reinstalls,  tried using DDU Display Driver Uninstaller in safe mode,  reinstalled Win 11,   upgraded to Win 11 22H2.  Same problem.

Turns out that having the GeForce Experience overlay enabled was the culprit.  

From now on when I find a good driver revision I'm going to keep a local copy of it.  

 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...