.

Epower, I've only run about a half dozen QCs while setting up the joystick, pedals, and TrackIR. I do have all those back to how I like them now, and I did run a six-on-six dogfight in late 1918 with all the eye candy, landscape density, and clouds maxed out, and it was B-E-A-U-tiful! Almost no preload required, and silky smooth panning with the TrackIR. And only one hiccup/micro-stutter, about ten seconds into the fight.

kk, thanks for the detailed outline on the two approaches to overclocking. I believe I will continue to let the Asus AI Suite take care of it for me.

.

Yeah, that's the thing; it wasn't always like that. Used to be, I could guarantee a big performance increase by fiddling around with cores, multipliers, voltages, etc...and it worked, sometimes getting 20% or more with suitable cooling. I was more than a little surprised when I saw a stock setup get a respectable increase by using 'factory' overclock settings.

I would have to say, though, that the factory overclocks seem to tend toward higher voltages than what are strictly necessary for the speed bump they deliver. This may be very different on later boards, though...I've only done up through Z390, and that was definitely me trying everything to get as high an all core overclock as I could (and it worked; got to 5.2 with a -2 AVX offset). Obviously, it depends a lot on your testing protocol. A lot of people don't use Prime95 anymore, because they feel it's just too much...me, I want to try to make sure overclocks are rock solid, so I usually shoot for 24 hours of P95, with AVX loads, in a closed case, at 80F ambient. I do this knowing full well the machines will *never* see this kind of load outside the shop, so if they're stable like that, they're aren't going to have problems with much less demands put on them. Of course, the downside is I could probably get even higher overclocks if the test were less demanding.

Anyhow, it is truly great to see how that thing's doing

.

Here's an update for those interested. With the new system parts pretty much dialed in the way I want them, I ran a baseline test with WOFF PE. 10v10 Se5s and DRVIIs directly over the front in November of 1918 with everything maxed out. Out of the box settings for TrackIR, default profile, no core affinities set; OOTB settings in the Nvidia CP; no W10 streamlining done at all. Repeated the test flight three times with consistent results as follows:

I had beautifully smooth tracking no matter how fast I turned my head. Saw a single white triangle on the edge of the screen for a split second, twice in the 1st run, none in the 2nd, once in the 3rd. In the 1st run I had a very brief hiccup/micro-stutter about 10 to 12 seconds in that lasted about a half-a-second and another of equal length late in the fight just above the mud when countless guns were firing at the three of us still in the go-round. 2nd and 3rd runs also each had the initial hiccup/micro-stutter, but no others that I noticed, (though to be fair fighting was quite intense so I may have missed them). There was a brief moment in the second test where it almost felt like everything slowed up slightly for a second or two, but not really sure about that one.

Overall, the new system kicks some serious flight sim backside with power and resources to spare, and that is before any of the usual WOFF tweaks. I've been reading that the "traditional" W10 streamlining for gaming probably isn't even necessary with the late generation cards, mobos, and CPUs, so I may forgo that entirely. The in-sim results right now are so outstanding I really don't think I'll be messing with much other than my custom profile for TrackIR.

Happy Happy Joy Joy!

.

.

Oh yes, I should also include the new system's full specs as well:

CPU: Intel Core i5-10600K Comet Lake, OC’d through the AI Suite to 5.0GHz
CPU Fan: Arctic Super Cooler
Memory: Corsair Vengeance 32gb (2 x 16gb) LPX DDR4 3200
Mobo: ASUS Prime Z490-A
Hard Drives: 2 Western Digital BLACK 500GB SN750 M.2 SSDs ; 1 Western Digital 640 GB Caviar Black SATA for onboard file storage
Opti Drive: LG 22X DVD+/RW Dual Layer SATA Rewrite
Video Cards: 2 EVGA RTX 2070 Super FTW3 Ultras with NVLink SLI Bridge
PS: Corsair HX 850 Watt
OS: Windows 10 Pro 64-bit
LG 34-Inch 21:9 curved IPS monitor with G-SYNC, 2560x1080 native resolution
Five large case fans, plus the PS, CPU, mobo, and card fans, (for a grand total of fourteen fans)
Saitek AV8R joystick
Saitek Pro Flight rudder pedals
Track IR4 camera with latest IR5 software

I have the OS and other assorted programs on the C drive, and WOFF PE only on the D drive. Everything else in on either the old SATA drive or the 1TB external drive I have plugged in.

.

Glad to see you're driving a monitor on the same level as your system.

Now for your next project - simpit! Only then will your time machine will be complete.

.

hee hee, the simpit may have to wait a while Epower. My total upgrade costs for the above system were as follows:

$525 for the CPU and mobo
$120 for the RAM
$170 for the SSDs
$800 for the GPUs and bridge
$80 for incidentals

For a grand total of $1,695 (prices do include tax and shipping). Now, while this did not entirely strip out my computer maintenance/flight simming improvement fund, it did put a VERY significant dent in it.


.
.

Lou, did you turn off the swap file completely with the 32 GB of RAM? Should save you some SSD life. That upgrade is free!

.

Fullofit, from what I've read recently it is better to just leave it all as Windows 10 64-bit has it. Things will likely run faster, and the impact on SSD life is negligible at best, if indeed affected at all.

.

Lou, if computers were governed by laws of physics and by logic, one would think that “things” running within the super fast RAM would run faster/smoother rather than swapping out to slower, albeit still fast, SSD. But since when one ever used logic to deal with computers?

.

What is this logic you speak of?

.

No matter how much RAM a Windows PC has, the swap file (or 'paging file', I believe) is used for when the tasks running exceed memory space; writing data from RAM out to a physical drive, then 'swapping' back and forth between the swap file space and RAM space whatever specific data is needed, and to create/maintain whatever space is deemed appropriate (by the OS).

If a PC has 32G of RAM, then it would obviously do this 'swapping' less often than a machine with 16G (given the same task load) , and a 16G machine less than an 8G setup, and so on.

Turning this swapping off means that, once you hit the limits of RAM, you can't use this functionality to help increase limits of what you can run. Turning it off doesn't mean it will run faster by being left in RAM, since normally it only 'swaps' when space is needed anyway - meaning the OS is set up to use faster memory by default, up to the limit it can. If you turn off swapping, you're still going to have the exact same RAM space limitation, only without having any place to go after that.

So, it would seem that you'd be setting more of a limitation and at the same time losing some functionality.

Or at least that's how it seems to me

Now, people may say that you can turn it off to save write activity to an SSD (reads don't matter)...and that's true - to a point. It is undeniably true that less writes to SSDs (more specifically, less erase and write cycles) mean the cells should 'live' longer. But I would assert that, these days, the technology has come so far with wear-leveling and such plus improvement in the durability of the NAND media, that you're not likely to see any ill effect from using an SSD as a boot drive with a paging file on it. Even though there is definitely wear on some scale, the drives are typically 'over-provisioned' around 7% if memory serves, so that as cells are retired due to write cycles, they are replaced by brand-new unused cells from that over-provisioning pool.

It is my understanding that, in typical desktop usage, these factors combined mean drives can go 10 years without suffering significant effect, and even longer - by which time (I believe) the assumption is you've long since moved on to another device.

Lou, so much knowledge to be gained by your experience...

Like not requiring change of core affinity for TrackIR - which is consistent with my dialogue thus far with NaturalPoint. And, in spite of a clearly monstrous setup, the familiar small issues inherent in the sim will still present themselves at times. The apparent confirmation that, short of maybe 4k resolutions, 3080 cards aren't going to double performance (and this is one area where I'd like to see more tests on your setup)...

But, as fascinating as all this is, I wonder what you could tell us about the sim's air activity settings. If you've mentioned it already, I apologize for missing it. But it happens, at this very moment, there's a separate thread with a discussion about Ultra Air activity. I am very curious if you've tried this setting (since you now have the world's first freakin' Cray SuperComputer Home Edition ) , and what your findings might be.

Lou, I hear good things about this logic thing. You should look into it.
KK, I’m not an expert, I’m not even a doctor, so I could be wrong, but I don’t think page file and swap file are the same thing. You are right, you shouldn’t turn off page file. I don’t even know if you can, as it controls memory allocation. Swap file, on the other hand, with 32 GB of RAM can comfortably be turned off for typical suite of programs including WoFF. There is little chance of running out of memory to need a swap file. Been running without a swap file for years and never have I ran out of RAM. Lou might be running (insert a program that gobbles RAM here) and he may prefer to keep the swap file on to avoid the limitation. It’s up to every individual to decide if they need that belt to go with their suspenders.

Well, I guess there's some confusion here, so let's try to correct that: As long as I've been fiddling with this stuff, the term 'swap file' (which actually applies to older versions of Windows, I believe, like NT) and 'paging file' (newer versions of Windows) have been used interchangeably. In fact, you can Google it now and find many references online to their being the same thing: >LINK<

Now, to be totally accurate, Microsoft did add a specific swap file (swapfile.sys) for Metro apps to Windows 8 and 10. This is discussed here: >LINK<

Read that at your leisure, but the important take-away is that this 'swap file' has absolutely nothing to do with so-called' virtual memory', which is what the paging file is for: simulating additional memory space to allow for increased overhead, as discussed above. The new "swap file", by comparison, is described as follows:

"In summary, the swapfile — swapfile.sys — is currently used for swapping out Microsoft’s new style of app. Microsoft has called these universal apps, Windows Store apps, Metro apps, Modern apps, Windows 8 apps, Windows 8-style UI apps, and other things at various points."

Which, I'm sure you can see, is not the same thing as Virtual memory. So, in the context of our performance discussion, I would submit it isn't relevant.

If we're being totally accurate, what you said was:

(my emphasis)

If it's inaccurate to use paging file and swap file interchangeably, it's equally inaccurate to use 'swap file' when I believe you're referring to the functionality of the paging file.

As far as changing the paging file (and corroboration that it is, in fact, about virtual memory), see here: >LINK< Note that this article also discusses the myth regarding disabling the paging file, in the same terms I described it above.

I do regret and apologize if my using the terms interchangeably causes confusion - however, please take into account, there are a lot of people out there who do this, and I'm still not altogether sure it's wrong. The paging file is often called a swap file (because that's what it was called previously), and both perform the management of Windows' "Virtual Memory".

Sounds like Microsoft has screwed decades worth of industry convention by deciding to create a new type 'swap file'...to me. I would venture that the overwhelming majority of experienced people - especially those mature enough to remember earlier versions of Windows - would say a paging file and a swap file are the same thing. I could be wrong, though.

Guess I'll need to stop using them interchangeably...learn something new every day. I appreciate the chance to learn.

I do hope this helps address any confusion.

.

Well I assumed we were talking about the "paging file" in this discussion of the "swap file", and my previous response was based on that assumption. In years past I have made the typical paging/swap file modifications and have seen varying levels of performance boosts in doing so. But as I said earlier, from what I've been reading there does not seem to be much benefit from doing it anymore. To be fair, if I were seeing any shortcomings in performance with my new set-up I probably would try the same tweaks to said paging/swap file that I've tried in the past. But right now I just don't think I will. The suspenders are working fine all on their own.

.

Lou, your conclusion is supported in my own research; that last link I included above includes references to testing and conclusions. I have also fiddled with paging file sizes/locations etc over the years and have ultimately concluded as you have.

So there we have it. There is no logic when it comes to computers, despite there being logical drives.
Now I'm going to be selfish and ask a question of my own. Should this setting be on or off, or (logically) it makes no difference to WoFF?

.

Don't know about others here, but I have it switched on. That being said, I tried it on and off with my old system and never noticed a difference. Still, it is an enticing switch.

.

You mean ... this enticing?


I didn't even know we have this button until now.

.

yes Stimpy, that enticing

.