1 hour ago, ryanbatc said:

I ran FT's Boston with FSLTL - the GPU/CPU was alternating a lot so it was being equally taxed I guess.

ideal. equilibrium. no need to upgrade anything then. 😊

16 hours ago, Fielder said:

Are you using Nvidia Control Panel / Configure Surround, PhsX to set up your 3 screens? Or the Experimental Multi Monitor in SU 10?

I find the PhysX Surround infinitely easier to set up. But both methods give better pixel density than just dragging the MSFS across 3 screens. 

I use the surround which gives me 5760x1080 on 4K 65" screens being driven as 1080p x 3. I tried the experimental multi-monitor and the performance is about 25% less than using the surround, but you get better side views. I'm sticking with surround until the 4090 and see if that can go 4K x 3 or at least center 4K and sides at 1080P.

1 hour ago, altenae said:

Is this a FSTLT topic or "Who is buying a 4090" 😉

"Bears,  Beets, Battlestar Galactica."  😁

https://www.youtube.com/watch?v=WaaANll8h18

Are the founders edition cards usually available directly from nvidia only?

Founders Edition are sold through Best Buy.

17 minutes ago, Waldo Pepper said:

Founders Edition are sold through Best Buy.

They got a 3090 Ti for 1100 USD.  I'm not sure how that will compare (power and performance) vs a 4080 16GB.

https://www.bestbuy.com/site/nvidia-geforce-rtx-3090-ti-titanium-and-black/6502626.p?skuId=6502626

15 hours ago, Beagle12 said:

Because MS2020 is CPU bound, I plan on upgrading to AMD Ryzen 9 - 7950X as soon as they become available here in Auckland NZ 🙂

MSFS seems to benefit from a massive L3 cache like the 5800x3D has.   I noticed the 7950X has about a one-third smaller L3 cache than does the 5800x3d....are you SURE the 7950X will outperform it in the sim?   I know if may bench out better, but I'm really interested in the in-sim performance.

I just wanna know this kind of information so I'm well-armed about exactly what cpu to go for when I upgrade next.

7 minutes ago, Mace said:

MSFS seems to benefit from a massive L3 cache like the 5800x3D has.   I noticed the 7950X has about a one-third smaller L3 cache than does the 5800x3d....are you SURE the 7950X will outperform it in the sim?   I know if may bench out better, but I'm really interested in the in-sim performance.

I just wanna know this kind of information so I'm well-armed about exactly what cpu to go for when I upgrade next.

I've mentioned it in the previous post, but going from 7700k to 5800X3D was a total game changer. I am glad I did not go for 12700 or 900. The gains are incredible. 

Just now, Drumcode said:

I've mentioned it in the previous post, but going from 7700k to 5800X3D was a total game changer. I am glad I did not go for 12700 or 900. The gains are incredible. 

That's why I'm wondering if the Ryzen 9 7950X will really truly beat the 5800X3D in the sim.

I know about that thread "5800X3D -- Brilliant Performance" for a reason 😆

Press play for MSFS with 12900k vs 7950x.    1080p  - 3090ti

 Of course,  this might be a stock install with no mods.   

I for one, have a lot of cpu intensive mods in my installation.  

Edited October 7, 20223 yr by Waldo Pepper

6 minutes ago, Mace said:

That's why I'm wondering if the Ryzen 9 7950X will really truly beat the 5800X3D in the sim.

I know about that thread "5800X3D -- Brilliant Performance" for a reason 😆

From what I've seen,  the 7950 doesn't beat a 5800x3d in msfs.   Someone posted benchmarks somewhere.

2 hours ago, Khashoggi said:

I use the surround which gives me 5760x1080 on 4K 65" screens being driven as 1080p x 3. I tried the experimental multi-monitor and the performance is about 25% less than using the surround, but you get better side views. I'm sticking with surround until the 4090 and see if that can go 4K x 3 or at least center 4K and sides at 1080P.

I use surround because I can't configure the experimental. When experimental  looks right, it doesn't stay looking right  when I pan around or translate around with the view hat. It all goes cockeyed. 

There's some videos supposedly showing how to use experimental but they skip so many steps, they aren't helpful.

I figured out on my own how to do some complicated control assignments (e.g. use a button to change speeds on yaw and pitch trim back and forth from fast trim to slow trim). But I can't figure out how to make the experimental multi screens look sensible when panning or translating the cockpit or external views. 

EDIT: Surround multi screens always look correct no matter how I move my viewpoint around. 

Edited October 7, 20223 yr by Fielder

Here's my problem. Using surround, thankfully it kills the view you used to have on one screen. Completely gets rid of it. But it also creates two new views on 2 screens: one view is the left half of the old image and the other view is the right half. And they move in unison no matter how you pan or translate the views. Perfect!

And then if you add another screen with surround you still do not have the old useless original view. Instead you have 3 new views left center and right. Synchronized. Perfect!

With experimental if I create 1,2, or 3 new windows, then I still have the old regular full cockpit view which is useless. And it hogs fps while doing nothing useful. People wind up with more new windows then then they originally planned on. Which kills fps.  And the old original window does not stay synchronized with the new windows when I pan or translate.

23 minutes ago, Fielder said:

Here's my problem. Using surround, thankfully it kills the view you used to have on one screen. Completely gets rid of it. But it also creates two new views on 2 screens: one view is the left half of the old image and the other view is the right half. And they move in unison no matter how you pan or translate the views. Perfect!

And then if you add another screen with surround you still do not have the old useless original view. Instead you have 3 new views left center and right. Synchronized. Perfect!

With experimental if I create 1,2, or 3 new windows, then I still have the old regular full cockpit view which is useless. And it hogs fps while doing nothing useful. People wind up with more new windows then then they originally planned on. Which kills fps.  And the old original window does not stay synchronized with the new windows when I pan or translate.

When I tested it out this is how I configured it-

The main screen is the main screen, so you only need to add two new windows through the experimental window interface. I set the main screen as 4K and also as 1080 to test both configs. I set the two new windows to windowed mode so I could drag them to the correct monitors and then set them to full screen so they used the whole monitor.

Started the sim at an airport I'm very familiar with, and then went back into settings and experimental and changed the roll, lateral and vertical axis of each monitor to give more view to the left and right. It worked well, but I went from a smooth 30FPS to around 22FPS or so and it just wasn't worth that loss. My Zotac 1080ti has served me well, but it is in a new 12900K build and it is time to make the leap to the 4 series GPU's and get up to 60FPS or 4K on all three at 30FPS which if I try now it is around 8FPS on the 1080ti.

3 hours ago, ryanbatc said:

They got a 3090 Ti for 1100 USD.  I'm not sure how that will compare (power and performance) vs a 4080 16GB.

https://www.bestbuy.com/site/nvidia-geforce-rtx-3090-ti-titanium-and-black/6502626.p?skuId=6502626

Newegg has the ASUS Tuf 3090ti for 1099.99 -- https://www.newegg.com/asus-geforce-rtx-3090-ti-tuf-rtx3090ti-o24g-gaming/p/N82E16814126555?Description=3090 ti&cm_re=3090_ti-_-14-126-555-_-Product -- unfortunately it is backordered now.

Edited October 7, 20223 yr by pgde