4 minutes ago, tfm said:

In Windows Task Manager, had anyone noticed the descriptions that crop up when you hover your mouse pointer over different segments of the informative bar chart? 

They imply that there is a use, albeit only as "standby", for larger amounts of RAM: in my case, with the amount of RAM showing as "Free" being c22Gb. On a 64Gb system, that obviously indicates that more RAM is being used than would be possible on a 32Gb system, albeit that this says nothing about whether "being used" is also "being used productively".

 

Screenshots here if interested:

https://share.icloud.com/photos/037eEARTSV6rD-xGxZDQhUreg

If you are talking about Standby RAM, then it is effectively cached data waiting for working sets (processes) to call it. Your images show you have 19.4 of Cached/Standby RAM. However, if another application starts up, and there is not enough free RAM to run it, then Standby/Cached will be dumped in favour of the new application. Standby memory is not really in use in the sense it is just storing low priority data however, because it can be dumped in a heartbeat, it is considered, along with Free memory to be your  'Available Memory'.  

Interestingly, if MSFS just is used to just sit and  plane spot, or on a long overwater flight,  the cached data will build up until there is no free memory available at all.  It all becomes standby. This just gives the Memory Manager more room to store data that can be quickly called if needed or just as quickly dumped and turned back into free memory for a new process.

 

Cheers

Terry  

16 hours ago, Bob Scott said:

@Lord Farringdon It's been a long while since I delved into this topic, and it was on a previous version of a MS flight sim, but I think the same dynamics likely apply here--namely, that the sims all do a pretty good job of lookahead caching, where approaching scenery is pre-loaded from nonvolatile storage (SSD/HDD) into memory, so any hard faults between RAM and the page file are occurring in the background well in advance of the scenery data actually being used, and thus aren't impacting the render threads.  I think we concluded, back in the day, that the big stutter experienced when coming into range of a complex airport was mostly attributable to the massive slug of 3D object data being transferred from RAM to the GPU when the in-range "switch" gets thrown to start rendering the airport and all of the myriad 3D objects associated with it.

FWIW, I haven't seen those big "approaching the field check" stutters in MSFS for quite some time now. 

I think you're right there Bob. I have seen hard faults at times of scenery loading but then there are a lot of other processes going on as well so it is hard to isolate what is actually causing the hard faults. It may not even be MSFS at all!! Having said that, I have tested this with FSLTL  and with a low frequency airport was definitely able to correlate hard page faults with spawned and approaching aircraft as LOD increased. However, These have not necessarily caused any stutter. With regards to Airport scenery,  as you say, it is indeed more likely that stutter is to do with  CPU/GPU interface which is another ball game altogether! 

Cheers

Terry 

10 hours ago, Noel said:

Finally got back to try KSLC -> KDEN in the 738.  Now nearly at cruise, this is late dusk to night flight, I have to say there was as much occasional micro stutter as w/ 32Gb.  After 38 minutes into this flight still no page file in use FWIW.  Changing views is instantaneous and perfect.  An acid test for me will be in the WT 787X which is hard on panning/view change compared to 738.  Is the consensus that 2024 will be more memory intensive?

Thanks for the update Noel. As I surmised above, it seems the extra memory has not removed the microstutters but it has reduced your paging requirements. However, I would like to see your Task Manager and Paged Pool though because @psolk  shows about 1.6GB in his paged pool on 64GB. I think yours will show something similar rather than being zero. 

The fact you are now getting instantaneous and perfect view changes is interesting and this just confirms what I said about higher speed memory  helping to support faster CPU/GPU throughput processing. Indeed just 32GB of DDR 5 may have done the same thing? All that said, 64GB is not about MSFS 2020 which you and I both know otherwise runs fine on 32GB but rather what the alpha testers say about the need for 64GB in MSFS 2024 and we still don't know that. 

Noel and @Cpt_Piett. Further to Performance Monitor and the lacking of paging detail. We have all been looking at the wrong thing. We need to add the counter 'Memory', then  select Page Faults/sec which returns soft and hard page faults combined, and also select  'Page Reads/sec' which seems to mostly correlate with hard page faults you will see in Resource Monitor. Both counters should read about 20 or below on average.  Given that MSFS2020 runs ok on 32GB with plenty of headroom and with oodles more headroom on 64GB, I doubt that the amount paging will be causing any concerns except in hardware limited systems, faulty memory or maxed out memory situations. 

 

Cheers

Terry

  

33 minutes ago, Lord Farringdon said:

Thanks for the update Noel. As I surmised above, it seems the extra memory has not removed the microstutters but it has reduced your paging requirements. However, I would like to see your Task Manager and Paged Pool though because @psolk  shows about 1.6GB in his paged pool on 64GB. I think yours will show something similar rather than being zero. 

Well, seems a bit early to establish that...unless ALL MICROSTUTTERS are related to memory management and I have to assume that's not the case--after all I have a grand total of one flight last night.  I couldn't see an entry for Paged Pool so if I need to look elsewhere let me know:

33 minutes ago, Lord Farringdon said:

The fact you are now getting instantaneous and perfect view changes is interesting and this just confirms what I said about higher speed memory  helping to support faster CPU/GPU throughput processing. Indeed just 32GB of DDR 5 may have done the same thing?

I have the same speed memory in the 64Gb kit as w/ the former 32Gb kit.

Sidenote re microstuttering:  over the years since MSFS released the most repeatable action to restore best performance is to flush the DX shader cache and update or reinstall the GPU driver doing a clean install.  Very often after this I'll have the most stutter-free performance and sometimes virtually completely free.   This morning's flight with intense scrutiny saw 3 ultra subtle but there microstutters during taxi then after that and during TO not a hint.  

 

 

 

Edited October 13, 20241 yr by Noel

 

wow city stutters with the 32GB above. Interesting.... not much different from the current sim issue(s) :)

2 hours ago, Sethos said:

 

So basically the same situation as MSFS2020.

7 hours ago, Lord Farringdon said:

I would like to see your Task Manager and Paged Pool though because @psolk  shows about 1.6GB in his paged pool on 64GB. I think yours will show something similar rather than being zero. 

I found it, it's 1.3GB.  My understanding is In Windows 11, the paged pool is a part of kernel memory that can be paged to disk (i.e., pagefile.sys) when physical RAM is running low or needed for other tasks. However, under normal circumstances, the paged pool remains in RAM until memory pressure requires it to be moved to the pagefile.  IOW, doesn't seem to come into play anyway with so much physical RAM overhead.  And, is somewhat corroborated by the 0% page file in use we see.

Edited October 13, 20241 yr by Noel

Ok well we now have 5 flights under our belt in a couple different scenarios and here is my take on 64Gb v 32Gb of the same decent DDR5 RAM.  Of note did the DX shader cache flush and installed the latest GPU driver:

Objective Findings:

• Eliminates paging to pagefile.sys, the value of which at this point is unknown to me.
• Increases physical RAM in use by roughly 20%
• Unable to fully verify CPU average temps are lower but they appear to be so by maybe 4-5 degrees C.  This is in part obscured because ambient is a bit cooler.

Subjective:

• Very hard to quantify but overall sim seems 'more responsive'.  This is clearly in the domain of potential placebo effect as I see it.
• Panning and changing views seems better, but not by a big margin by any stretch as it was already excellent.
• Micro-stuttering and overall sim "smoothness" is largely equivocal.  Prior to installing 64Gb I had been thinking that overall performance has been really quite good for a while now.  I had some stutters taxiing at a very modest airport (KOMA) upon arrival and was surprised to see that.  And yet for flight #5 out of KEWR didn't see a one and that is dramatically more intense than KOMA was.  So in this area which I identified early on as the main purpose for upgrading I would have to honestly say in a double blinded test I'm not sure I'd choose 64Gb more than 50% of the time in multiple tests as other factors are likely impacting micro-stuttering.   It's still early to have a strong opinion on this aspect.

Bottom line:  I'll keep the 64Gb as it's having some impact and the cost was modest relative to total system cost.

Cheers

 

 

32 minutes ago, Noel said:

Ok well we now have 5 flights under our belt in a couple different scenarios and here is my take on 64Gb v 32Gb of the same decent DDR5 RAM.  Of note did the DX shader cache flush and installed the latest GPU driver:

Objective Findings:

• Eliminates paging to pagefile.sys, the value of which at this point is unknown to me.
• Increases physical RAM in use by roughly 20%
• Unable to fully verify CPU average temps are lower but they appear to be so by maybe 4-5 degrees C.  This is in part obscured because ambient is a bit cooler.

Subjective:

• Very hard to quantify but overall sim seems 'more responsive'.  This is clearly in the domain of potential placebo effect as I see it.
• Panning and changing views seems better, but not by a big margin by any stretch as it was already excellent.
• Micro-stuttering and overall sim "smoothness" is largely equivocal.  Prior to installing 64Gb I had been thinking that overall performance has been really quite good for a while now.  I had some stutters taxiing at a very modest airport (KOMA) upon arrival and was surprised to see that.  And yet for flight #5 out of KEWR didn't see a one and that is dramatically more intense than KOMA was.  So in this area which I identified early on as the main purpose for upgrading I would have to honestly say in a double blinded test I'm not sure I'd choose 64Gb more than 50% of the time in multiple tests as other factors are likely impacting micro-stuttering.   It's still early to have a strong opinion on this aspect.

Bottom line:  I'll keep the 64Gb as it's having some impact and the cost was modest relative to total system cost.

Cheers

 

 

Your conclusions are the same as mine when it comes to MSFS 2020.
But I've seen a video from a users who tried the Alpha version of MSFS 2024 saying that 2024 uses significantly more RAM than 2020 does, https://youtu.be/4RjAFaJY3bI?t=1482 almost 32GB in ultra settings, so hopefully it wasn't a waste of money after all.

3 hours ago, Ixoye said:

so hopefully it wasn't a waste of money after all.

I'm masterful at rationalizing just ask my spouse: 

• Hardware including home built PC this year, LG Ultragear G-sync Ultimate 34" curved + TM Boeing Yoke & TQ 3yr ago or so around $4,700.
• MSFS w/ various addons around $700.

64Gb@$215/($4,700+$700) = just under 4% of total cost, so not a big investment as has been said by others.

Ahh yes. My standard line to my wife when asking about new hardware. “Oh that. I’ve always had that”

On 10/13/2024 at 12:11 PM, Sethos said:

 

@ryanbatc this video may be of interest to you. The panning stutter seems to be eliminated with 64GB.

Got 2x32GB installed. Luckily I was able to use a BIOS profile I had for 32GB, so primary and secondary timings are identical.

I’ll run TestMem5 for a while, then boot up the sim and compare performance with 32GB. I’ve already tested ILS approaches into KLAX and EGLL with 32GB. Should be interesting to compare CapFrameX tests.

I did 4 tests with exactly the same settings, as well as active applications:

1. Fenix ILS approach into KLAX
2. Fenix ILS approach into EGLL

I was surprised by the findings, to be honest! 

In both tests, average FPS as well as avg. 1% and 0.1% lows were significantly higher with 64GB. In addition, the frame time variance was significantly better with 64GB, indicating that the sim ran smoother. 

MSFS as well as total RAM usage was also higher with 64GB. VRAM, CPU and GPU usage did not change significantly. 

...

I'll post the numbers here in case anyone's interested:

KLAX 32GB => 64GB

Avg. FPS 66.5 => 69.3

1% low avg. 22.4 => 27.5

0.1% low avg. 13.6 => 18.6

FTV <2ms 62.7% => 69.4%

MSFS/total RAM usage 6.3/23.9GB => 11.8/28.6GB

EGLL 32GB => 64GB

Avg. FPS 64.8 => 73.5

1% low avg. 22.8 => 25.8

0.1% low avg. 12.5 => 16

FTV <2ms 63.8% => 67.9%

MSFS/total RAM usage 6.6/23.8GB => 11.1/27.8GB

...

In both tests I used 4K, DLSS+DLAA, FG, TLOD 50, OLOD 200, ultra settings. AIG TG at 50%. 

...

I'll publish a video on my channel in a few hours. 

Edited October 14, 20241 yr by Cpt_Piett