Ryzen 7 5800X3D - Brilliant MSFS performance

[Maxis 1,857]

16 minutes ago, Makinen11 said:

. 

Saying that, we know that with DLSS nvidia is the way to go for MSFS no matter what. 

That's debatable.. The other shoe hasn't dropped yet for FSR 2.0 implementation. 

So another wait and see approach for that aspect as well.

Edited August 1, 2022 by Maxis

[Greazer 1,320]

11 hours ago, ca_metal said:

That’s impossible. The socket types are different. AM4 is PGA and the 7000 series will be LGA.

Not "impossible". The 7000 series may support both.

https://www.pcgamesn.com/amd/ryzen-7000-zen-4-cpus-ddr4-support

AM5 motherboards will only support DDR5 gaming RAM, but some Ryzen 7000 processors may be destined for the AM4 platform which can only use DDR4 sticks

AM4 is a very popular platform. AMD will be wise to support it. Gamers can upgrade just their CPU and not the whole PC!  That's how it is supposed to work.

[zecanelao 1]

Are you guys using hyperthreading?

[Virtual-Chris 682]

22 minutes ago, zecanelao said:

Are you guys using hyperthreading?

Yes. AMDs SMT in my case. 

[Maxis 1,857]

31 minutes ago, zecanelao said:

Are you guys using hyperthreading?

Yes . SMT as well

[Kanade 15]

I did the benchmarks focusing on MSFS2020 (under generally standardized conditions) on a processor (5950X/5800X3D) with almost no differences other than the cache to see the effect of the cache.

The discussion can be found in the official MSFS2020 forum.
FPS difference verification between AMD 5950X and 5800X3D at Haneda Airport (RJTT) 34L straight out flight

Two of the main charts are also described here.

 

[espent 1,390]

1 hour ago, Kanade said:

I did the benchmarks focusing on MSFS2020 (under generally standardized conditions) on a processor (5950X/5800X3D) with almost no differences other than the cache to see the effect of the cache.

The discussion can be found in the official MSFS2020 forum.
FPS difference verification between AMD 5950X and 5800X3D at Haneda Airport (RJTT) 34L straight out flight

Two of the main charts are also described here.

Nice analysis. I am surprised to see the huge difference between the 5950 and the X3D. On the fence if I should jump or wait for next gen.

I feel I am not getting full value for my 3090 atm, so this might be the trick to fix it.

[ca_metal 2,015]

22 hours ago, Greazer said:

Not "impossible". The 7000 series may support both.

https://www.pcgamesn.com/amd/ryzen-7000-zen-4-cpus-ddr4-support

AM5 motherboards will only support DDR5 gaming RAM, but some Ryzen 7000 processors may be destined for the AM4 platform which can only use DDR4 sticks

AM4 is a very popular platform. AMD will be wise to support it. Gamers can upgrade just their CPU and not the whole PC!  That's how it is supposed to work.

Ok, impossible wasn't the best word, sorry. 

What I mean is their Zen 4 was projected to the AM5 socket, to be used with ddr5, pci-ex 5.0 etc. Even, if they decide at some point to release a 7000 series compatible with AM4, I'm not so sure it will be a Zen 4 CPU (probably a rebranded Zen 3). 

I can see them releasing an AM4 "7400G". But I really doubt they would realease an AM4 "7950x". 

Being reallistic, the best options for the AM4 from now on is AMD releasing new X3D CPUs. Talking about performance.  

Edited August 2, 2022 by ca_metal

[Virtual-Chris 682]

2 hours ago, Kanade said:

I did the benchmarks focusing on MSFS2020 (under generally standardized conditions) on a processor (5950X/5800X3D) with almost no differences other than the cache to see the effect of the cache.

The discussion can be found in the official MSFS2020 forum.
FPS difference verification between AMD 5950X and 5800X3D at Haneda Airport (RJTT) 34L straight out flight

Two of the main charts are also described here.

 

It’s interesting that you found such a difference with SMT off. I will definitely be trying that myself. 

[Kanade 15]

I remember that the performance of the MSFS2020 was very confusing from the time it shipped, the ratios contributed by each component of the PC being very difficult to understand.
The correlation between the results of various benchmarks and the MSFS2020 (especially for FPS and responsiveness) was quite low, and I did not see anyone doing a satisfactory analysis, so I decided to get my own data.

MSFS2020 is (probably) not suited for conventional methods of the nature of being able to say "average nn FPS" throughout the entire game. The load fluctuations are so severe that averaging them out would miss the correct result.
Many users say that their CPU should be more capable, but MSFS2020 is not using it to its full potential.
Very true, but on the other hand it is not accurate.
It is not the ALU (execution part) that is missing, but the data supply part.
The execution part has no choice but to stop and wait for data to arrive.
Also, the bandwidth of data transfer is important, but even if the bandwidth is excellent, it does not mean that data can be supplied without delay.
Latency is more important among multi-processor, tightly coupled applications.

The 5800X3D produces a latency reduction of about 300% compared to other processors, "only with about 70MB more L3 cache area" compared to main memory. That is the only difference.
(Main memory is typically 70-100ns latency in modern systems, L3 is around 20ns. The ALU only waits about 1/3 of the time for a hit.)
I think MSFS2020 is a typical case of an application that happens to have a cache requirement of 32MB or more.
 

[Kanade 15]

The 5800X3D is only effective in a limited number of scenarios because it makes no difference in cache hit rate for applications optimized to fit in a typical cache size (L3 16-32MB). (especially most office and benchmark applications).

In the traditional scenario, the requirement for maximum processor utilization is controllable by the programmer, so performance metrics are monitored during the development phase and if the speed falls below a certain level, the number of objects is limited or all relevant parts are removed.
In such applications, the performance is determined by the strength of the processor itself, including the processor ALU and instruction decoder. This is the "more obvious and typical" case for the average PC user.

MSFS2020 utilizes machine learning to automatically generate vast terrains. Not all conditions are under programmer control. There are probably many untested scenarios.
(This is not due to lack of optimization, but to the fact that the content is growing to a size that is beyond human control.)

 

[Virtual-Chris 682]

It would follow that besides increasing CPU cache size and reducing latency, main memory bandwidth (and latency/timings) are another major consideration when architecting a system. Which is, again, unlike almost any other game. 

It would be interesting to see some testing on memory bandwidth and latency in MSFS. What feeds that L3 cache quickest probably wins!

Edited August 3, 2022 by Virtual-Chris

[Kanade 15]

The Zen architecture is generally more sensitive to memory bandwidth and latency than the Intel architecture. Therefore, we expect to observe a reasonable speed change, especially up to DDR4-3600. (For DDR4-3600 and above, it is implementation-dependent on the motherboard manufacturer, since the Infinity Fabric standard is 2:1 with memory and only supports up to 1800 MHz.)

However, in the current Ryzen5000 series CPUs, there are already frequent memory accesses overflowing the cache, and in the scenarios I have seen, it is difficult to hope for that much performance push-up, as some timing is already stuck at 142W, which is the PPT.
We could remove the PPT lock, but I have studied the performance/power of each Intel/AMD processor in the past, and I know from experience that the current Zen3 architecture's PPT of 142W is an excellent balance point, so removing it would not do much good We believe that it will not.

Aside.
In the past, we have repeatedly tested CineBench on many CPUs, in 10W increments.
The goal was to validate the processor manufacturer's PPT=142W.
We wanted to experience that processor manufacturers do not make these decisions at random.
Benchmarking and plotting multiple CPUs in 10W increments from 84W (EcoMode)-230W (SMT+AllCore Static 4.9GHz), we found that the default 142W was an excellent point.
 

[pstrub 998]

12 hours ago, Kanade said:

I did the benchmarks focusing on MSFS2020 (under generally standardized conditions) on a processor (5950X/5800X3D) with almost no differences other than the cache to see the effect of the cache.

The discussion can be found in the official MSFS2020 forum.
FPS difference verification between AMD 5950X and 5800X3D at Haneda Airport (RJTT) 34L straight out flight

[...]

Thank you very much for the great analysis effort. It's interesting to see the fps gains in certain situations that can be achieved from turning off SMT. I assume you disabled it from the BIOS setup?

This doesn't seem very practical in my case, as I run a variety of use cases on my PC, some of which can make good use of SMT. Some people have suggested to use Process Lasso for disabling the SMT cores for MSFS (I know it isn't exactly the same), and I think it would be great to have such an option in Ryzen Master (although this might not be possible with the windows scheduler). Do you have any experience with these options?

I won't be able to test on my simming rig for a couple of days, but will report back when I find the time to play around with these tools.

Edited August 3, 2022 by pstrub

[Alpine Scenery 1,609]

So its worth it but only if disabling SMT is the conclusion, I am assuming...