Benchmark scenario
Takeoff from (Aerosoft Heathrow) EGLL 9L, Climb straight to 3000, turn right and intercept ILS 27L, autoland). No weather. (P3Dv4 All sliders to Max)

Addons
P3Dv4.4 HF2, ORBX FTX Global, Aerosoft Heathrow, FSL a320X (Latest ver).

System Specs
1. 8700K (5.0ghz, AM 1365), Noctua NH-U14s, 32GB DDR4 3200 CL16 (4x8GB), RTX 2080, 1440p resolution.
2. 3800X (4.375ghz, HT Off), Stock cooler, 32GB DDR4 3200 CL16 (4x8GB), RTX 2080, 1440p resolution.

Results

8700K
                     Average framerate  :   38.1 FPS
                     Minimum framerate  :   31.6 FPS
                     Maximum framerate  :   42.3 FPS
                     1% low framerate   :   27.3 FPS
                     0.1% low framerate :   22.6 FPS

3800X
                     Average framerate  :   34.4 FPS
                     Minimum framerate  :   28.0 FPS
                     Maximum framerate  :   38.6 FPS
                     1% low framerate   :   26.7 FPS
                     0.1% low framerate :   24.1 FPS

Observations
3800X performs worst with SMT enabled (but still smoother than 8700k stock with no AM). It performed much better with AM 21845, but performed best with SMT disabled.
Also, disabling SMT brought down temps by 10c atleast. My 3800x system OC was limited by operating temps, the stock cooler couldnt handle a 4.5ghz all core OC, even though it was stable. Also, going from 3200c16 2x8GB to 4x8GB improved performance by atleast 5%.
With default PBO boost, 3800x clocks were hovering around 4.1-4.3ghz, but weren't steady, hence requiring an all core OC, which stabilized clocks.

Feel free to ask any questions and suggestions you have. I'll try my best to test it out in the sim.

Edited February 22, 20206 yr by asgnp

I would think that a far better comparison would be with both CPU's running at the same frequency.  Either overclock the AMD processor or limit the Intel processor.  Still, it is an interesting comparison and if increasing the CPU frequency is or is very near a linear increase in fps, then the AMD processor might be slightly better than the Intel processor.  At a minimum, they would be equals.  I find this to be...

Edited February 23, 20206 yr by stans

34 minutes ago, stans said:

I would think that a far better comparison would be with both CPU's running at the same frequency.  Either overclock the AMD processor or limit the Intel processor.  Still, it is an interesting comparison and if increasing the CPU frequency is or is very near a linear increase in fps, then the AMD processor might be slightly better than the Intel processor.  At a minimum, they would be equals.  I find this to be...

I tested the 3800x at various frequencies from 4 ghz to 4.4 ghz, the increase in fps is pretty much linear.

21 hours ago, asgnp said:

I tested the 3800x at various frequencies from 4 ghz to 4.4 ghz, the increase in fps is pretty much linear.

Quite fascinating.

On 2/22/2020 at 6:11 PM, asgnp said:

Benchmark scenario
Takeoff from (Aerosoft Heathrow) EGLL 9L, Climb straight to 3000, turn right and intercept ILS 27L, autoland). No weather. (P3Dv4 All sliders to Max)

Addons
P3Dv4.4 HF2, ORBX FTX Global, Aerosoft Heathrow, FSL a320X (Latest ver).

System Specs
1. 8700K (5.0ghz, AM 1365), Noctua NH-U14s, 32GB DDR4 3200 CL16 (4x8GB), RTX 2080, 1440p resolution.
2. 3800X (4.375ghz, HT Off), Stock cooler, 32GB DDR4 3200 CL16 (4x8GB), RTX 2080, 1440p resolution.

Results

8700K
                     Average framerate  :   38.1 FPS
                     Minimum framerate  :   31.6 FPS
                     Maximum framerate  :   42.3 FPS
                     1% low framerate   :   27.3 FPS
                     0.1% low framerate :   22.6 FPS

3800X
                     Average framerate  :   34.4 FPS
                     Minimum framerate  :   28.0 FPS
                     Maximum framerate  :   38.6 FPS
                     1% low framerate   :   26.7 FPS
                     0.1% low framerate :   24.1 FPS

Observations
3800X performs worst with SMT enabled (but still smoother than 8700k stock with no AM). It performed much better with AM 21845, but performed best with SMT disabled.
Also, disabling SMT brought down temps by 10c atleast. My 3800x system OC was limited by operating temps, the stock cooler couldnt handle a 4.5ghz all core OC, even though it was stable. Also, going from 3200c16 2x8GB to 4x8GB improved performance by atleast 5%.
With default PBO boost, 3800x clocks were hovering around 4.1-4.3ghz, but weren't steady, hence requiring an all core OC, which stabilized clocks.

Feel free to ask any questions and suggestions you have. I'll try my best to test it out in the sim.

8700K at 4.4 ghz
                     Average framerate  :   34.6 FPS
                     Minimum framerate  :   28.0 FPS
                     Maximum framerate  :   39.0 FPS
                     1% low framerate   :   24.5 FPS
                     0.1% low framerate :   18.6 FPS

So with all things being equal, including CPU frequency, one would find no appreciable difference in FSX performance between an Intel 8700K CPU and an AMD 3800X CPU?  This makes AMD a player in the flight sim market.  Competition is a good thing.

Edited February 29, 20206 yr by stans

It would be interesting to see what effect overclocking the AMD Infinity Fabric would have. Several YT videos seem to imply that would yield increased performance.

I think all flight sim users should now pill in on AMD then I will be able to buy my Intel CPU on the cheap at the end of the year.😁

2 hours ago, pgde said:

It would be interesting to see what effect overclocking the AMD Infinity Fabric would have. Several YT videos seem to imply that would yield increased performance.

On my system, overclocking fclk did not scale as well as cpu frequency... but 1% lows improved quite a bit.

6 hours ago, stans said:

So with all things being equal, including CPU frequency, one would find no appreciable difference in FSX performance between an Intel 8700K CPU and an AMD 3800X CPU?

That would certainly happen if one is GPU-limited. Not saying that's what happened here, but I'd love to see the GPU utilization.

Cheers!

It's all very well comparing the two processors at the same clock frequency, but if (for example) the i7 8700k has significantly greater overclocking potential, then it stands to reason that (all other things being equal) this processor would be the better choice for P3D (which is basically what the OP indicated in his original post).

Edited March 1, 20206 yr by Christopher Low

22 hours ago, Luke said:

That would certainly happen if one is GPU-limited. Not saying that's what happened here, but I'd love to see the GPU utilization.

Cheers!

Gpu utilization was 35-55%.

11 hours ago, Christopher Low said:

It's all very well comparing the two processors at the same clock frequency, but if (for example) the i7 8700k has significantly greater overclocking potential, then it stands to reason that (all other things being equal) this processor would be the better choice for P3D (which is basically what the OP indicated in his original post).

Yes, intel chips are still the optimal cpus for p3d solely because of their overclocking ability, if you’re after the last extra fps. More importantly what i wanted to show is that even under very resource intensive sim scenarios, amd chips like that 3800x will provide more than adequate performance to maintain 30fps and provide smooth 30fps locked gameplay.

Thank you for the head-to-head comparison of comparable AMD and Intel CPU's.

On 2/29/2020 at 3:15 PM, G-RFRY said:

I think all flight sim users should now pill in on AMD then I will be able to buy my Intel CPU on the cheap at the end of the year.😁

🤣