Hyperthreading support up to 16 cores in DX10 is possible!

[BiBa]

In my last thread there was the discussion which is better Intel or AMD CPU for IL-2 hyperthreading? The majority declared Intel as the winner due to threading limitation. Now again I'm not sure any more...

 

Recently I bumped into a utility Add-On which made me exhume the old FSX dinosaur out of its grave for an unbelievable revival and gave it wings with such a performance, I couldn't believe it was possible. All my six cores were in full operation to the extreme limit and boosted my frames to a hitherto impossible height! (see attached pic) This utility offers also many configuration options in selecting Add-On priorities and affinities for further boosting. 

 

So if hyperthreading support up to 16 cores is possible in DX9 + DX10 with such an old stiff engine like FSX, why is it not possible in a new flexible engine like IL-2?

Immagine the IL-2 performance with the AMD Ryzen Threadripper 1950X 16 cores all in use simultaneously!

 

Beyond that this greek team also offers an Add-On FSX Booster that can decrease / increase FSX frames WHILE playing, so that in a dog fight the Gamer can simultaneously increase the frames, or decrease them for a higher resolution to enjoy the beautiful environment sight. 

 

This is indeed the unavoidable challenge to the IL-2 programmers. 
Red-Bull does not give wings to IL-2. Those Add-Ons do

[SAS_Storebror]

I sense a great disturbance between game performance, multithreading, hyperthreading and the role of DirectX in all of this.

 

Mike

[BiBa]

2 hours ago, SAS_Storebror said:

I sense a great disturbance between game performance, multithreading, hyperthreading and the role of DirectX in all of this.

Mike

 

 

I'm aware of the difference between the two, nevertheless, this Multicore-App does feature Supporting Hyper-threading enabled systems. I am also aware that IL-2 has a different engine than FSX. But the fact is: this App did break the two cores limitation barrier in FSX! (which uses DX9 or DX10)

 

Now I assume IL-2 uses DX11. Here is a summary from the Internet on DX:

1) DirectX is an API (Application Programming Interface) used for developing video games and other such media applications, much like OpenGL. 
2) The differences between DX9 and DX10 is that Anti-Aliasing is supported in DX10 and DX10 supports shader v4.0 Anti-Aliasing is used to smooth the edges of objects in a video game and makes little difference in games playing in low resolutions. With it on, it will noticeably slow down performance in certain games while delivering an almost unnoticeable difference in visuals. The difference between DX10 and DX11 is that DX11 supports Tessellation. Hypothetically, DX11 is supposed to enhance performance with compatible systems as well as providing a graphics increase with a feature called tessellation. An example of the graphical improvements with tessellation on can be viewed here: http://www.youtube.com/watch?v=KtbUWiZ48a0 , tessellation is turned off at first. Too much tessellation can cause terrible lag just like too much anti-aliasing.

 

And a summary on Multithreading & Hyperthreading:

- Hyperthreading is a hardware thing and Intel branding. Most other people call it Simultaneous Multithreading (SMT). To the programmer, two hyperthreads look like two CPU cores. On the hardware side, multiple hyperthreads share a single core. (In the case of Intel, there are two hyperthreads per core).

- Multithreading (or multithreaded programming) is generally considered the concept of using more than one thread context (instruction pointer, registers, stack, etc.) in a single program. (Usually in the same process or virtual address space).

 

Let alone future Vulkan or DX11 apart, to frame the core of this topic, (in case I do decide to enhance my PC with a 16 cores AMD CPU) the question remains if IT IS possible to engage the remaining unemployed CPU cores turning thumbs and waiting for a job. In that case, I will decide for AMD and not Intel CPU.

 

if IL-2 is flooded to the neck with many tasks at the same time, I wonder if it is not wise to contract FSPS with that 3rd Party Multicore-App. If they were able to crack that Microsoft FSX engine, I'm sure with a IL-2 collaboration of your Program engine source code they'll find something. In case of failure, they'll be liable for the time investment damage themselves, and IL-2 has nothing to lose!
It is a win-win situation.

Edited August 27, 2018 by BiBa

[SAS_Storebror]

1 hour ago, BiBa said:

this App did break the two cores limitation barrier in FSX!

Fact is that FSX SP2 uses all physical cores of your system already.

 

1 hour ago, BiBa said:

To the programmer, two hyperthreads look like two CPU cores. On the hardware side, multiple hyperthreads share a single core.

Correct and as such, HyperThreading is not what you're asking for if you want to get better performance from a game.

It does have it's right to exist, but HT really isn't that much of importance for us IL-2 players.

 

1 hour ago, BiBa said:

if IT IS possible to engage the remaining unemployed CPU cores turning thumbs and waiting for a job. In that case, I will decide for AMD and not Intel CPU.

I'm afraid this won't be an easy thing to do for the devs.

Optimizing something for multicore use that hasn't been built for this purpose from the very beginning is an extremely tedious, sometimes pretty much impossible task.

 

Mike

[ZachariasX]

8 hours ago, SAS_Storebror said:

Fact is that FSX SP2 uses all physical cores of your system already.

True. It absolutely does. It doesn't do hyperthreading and it tends to have its main thread clogged much faster than current P3D v4.3 that had 10 years of some lovin' invested in that. Makes me wonder what the "greek software" would do.

 

[peregrine7]

19 hours ago, ZachariasX said:

True. It absolutely does. It doesn't do hyperthreading and it tends to have its main thread clogged much faster than current P3D v4.3 that had 10 years of some lovin' invested in that. Makes me wonder what the "greek software" would do.

 

 

10 bucks on absolutely jack all.

[Dutch2]

I have read some comment about this FSPS multicorenext software and to me it appears that it only sets the core affinity and prior. If like to experimentate you can use Windows own affinity and prior in the Windows task manager. 

If having succes you can always try free software like, 

https://wj32.org/processhacker/nightly.php

http://systemexplorer.net/

Which can set the affinity and prior adjustments permanent, without using to much recources like complex ProcessLasso. 

 

 

 

[ZachariasX]

26 minutes ago, Dutch2 said:

I have read some comment about this FSPS multicorenext software and to me it appears that it only sets the core affinity and prior. If like to experimentate you can use Windows own affinity and prior in the Windows task manager. 

If having succes you can always try free software like, 

https://wj32.org/processhacker/nightly.php

http://systemexplorer.net/

Which can set the affinity and prior adjustments permanent, without using to much recources like complex ProcessLasso. 

 

 

 

The problem is how windows didtributes processs among the cores. For flightsim, you‘d actually need to have a know, most overclockable core reserved to its main „fibre“ thread, while carefully distributing the other tasks among the cores. Windows is so bad at this that noone in high preformance computing woukd touch windows even with a stick.

 

Don‘t count on such software. Rememer „Ram Doubler“?

[SAS_Storebror]

1 hour ago, Dutch2 said:

I have read some comment about this FSPS multicorenext software and to me it appears that it only sets the core affinity and prior. If like to experimentate you can use Windows own affinity and prior in the Windows task manager. 

If having succes you can always try free software like

...FUMS, and make your life easier:

https://www.sas1946.com/main/index.php/topic,59488.0.html

 

Mike

[Dutch2]

I do not know if it does work for RoF/BoX, because RoF/BoX is already spreading its workload my CPU cores, in my case a i7-7700k quadcore, HT disabled,so it is not needed. But for an single core application like WoFF it was rather helpfull. I did reserve 3 cores exclusive for the .exe file, while one core was reserved for ALL the other activities, like OS, TrackIR etc. While being a single core application I did see spreading its load over two cores, while the 3th always was low on load. 

I must admit, in this specific single core game, I did use another program called CPU control, very lean and because very old, it is limited to 4core max, does not support HT and it does regonize not all the running processes on your computer. It does work on Win10-64 btw. http://www.koma-code.de/  while the other two from my first post do not have that limit and I think allso better suited. 

There is also Bill2s that seems to have the same functions, only I could never get this working. 

 

 

 

Edited September 23, 2018 by Dutch2

[SAS_Storebror]

The key feature of FUMS, and as much as I understand it's a unique feature among the other tools listed here, is that FUMS will automatically adjust your Windows Power Plan settings when a specific process is active (and reset it to your defaults when the process exits), so you can automatically follow the Dev suggestions for graphics smoothness given here without having to run your whole system in "Maximum Performance" mode all the time:

 

All the other features, like "process to core" assignments and priority settings, are covered by FUMS anyway.

Apart from that, FUMS is Open Source, so there's no highly confidential tool fiddeling with your system there.

 

Mike

[SCG_ErwinP]

On 8/26/2018 at 1:37 PM, BiBa said:

Immagine the IL-2 performance with the AMD Ryzen Threadripper 1950X 16 cores all in use simultaneously!

Developing multicore applications isn't so easy to do (I would love to see more threads been used, but...).

On 8/27/2018 at 5:00 AM, BiBa said:

I will decide for AMD and not Intel CPU

Not only does the number of cores matter, even though you're not using your computer just to play, I think.
Most game titles benefit from both high CPU clock and Instruction per cycle (IPC which is the volume of instructions the processor reads per clock cycle).
In this case, Intel processors have advantages in both things, that is, they have higher IPC and bigger clocks.
Users like you and I would only benefit from fully prepared applications for multithreads. Today, BOS uses only only 4 CPU cores.

On 8/27/2018 at 6:10 AM, SAS_Storebror said:

but HT really isn't that much of importance for us IL-2 players.

I agree, in MSI Afterburner it only shows 4 cores being used.

To use with VR, i think that 8 high clock cores is enough.

 

Criticisms and suggestions are welcome. 

[ZachariasX]

6 hours ago, SAS_Storebror said:

that FUMS will automatically adjust your Windows Power Plan settings when a specific process is active

So it is doing what Windows is supposed to do. If it was that easy, everyone does it. That things like that pretty mucgh don't work at all is nicely illustrated by the "Turbo Bosst max 3.0" by Intel. You can have it or not, your single thread performance is not gonna change at all for practical purposes, while it is still prone to produce thread congestions on the other hand. This means, it doesn't make your CPU faster, but you are more at risk for stutters.

 

If things don't work, adding complexity is the daftest thing you can possibly do.

[SAS_Storebror]

Probably the most uneducated reply I could have expected.

Thanks for mixing up each and everything involved and consequently drawing false conclusions.

You definitely have no clue what you're talking about.

Sorry if my reply sounds harsh, but yours is just so flippant, it's beyond imagination.

 

Regards 

Mike

[ZachariasX]

1 hour ago, SAS_Storebror said:

Probably the most uneducated reply I could have expected.

Thanks for mixing up each and everything involved and consequently drawing false conclusions.

You definitely have no clue what you're talking about.

Sorry if my reply sounds harsh, but yours is just so flippant, it's beyond imagination.

 

Regards 

Mike

That‘s good news then. How much improvement do you get from Inels Turbo boost?

[SAS_Storebror]

Who compared FUMS and Intel Turbo boost?

Hint: It wasn't me.

[ZachariasX]

30 minutes ago, SAS_Storebror said:

Who compared FUMS and Intel Turbo boost?

Hint: It wasn't me.

No, it came from someone who even used Stacker, RamDoubler (but just once, although that cost $50) and on top of that "optimized the registry of his computer" many times. You're talking to the bottom of the pit here.

 

So, a person like that, if he/she has a rig running at X fps, why do I install/run the software that "(expletive) up my system"? Do I get X+N FPS?

 

I mean, a person with an inclination to play around with the computer, there is certainly sympathy for the sheer principle of doing it. Is it that?

 

Who ever got a FPS increase by the use of "Process Lasso"? There are many people using it. Is there a game benchmark out there showing significant performance gains?

 

And why is it that one needs the Windows process scheduler what to do?

[SAS_Storebror]

Trying to blame it on others now?

You (and no one else) said about FUMS that it'd be "doing what Windows is supposed to do."

Why would Windows be supposed to set the Power Plan to "Maximum Performance" to follow Jason's advice about stutter-free gaming?

 

Mike

[ZachariasX]

15 minutes ago, SAS_Storebror said:

Trying to blame it on others now?

I was just offending myself and noone else. I was being sarcastic.

 

17 minutes ago, SAS_Storebror said:

Why would Windows be supposed to set the Power Plan to "Maximum Performance" to follow Jason's advice about stutter-free gaming?

You use the tool for doing that?

 

 

[SAS_Storebror]

Oh man...

If you have no idea what the tool is doing, how come you feel entitled to judge about it's purpose then?

Is it silly season again?

[ZachariasX]

47 minutes ago, SAS_Storebror said:

Oh man...

If you have no idea what the tool is doing, how come you feel entitled to judge about it's purpose then?

Is it silly season again?

No, let's be more serious here. Can you quantify performance gains using that tool? Or any of these tools?

[SAS_Storebror]

Did anyone claim such?

What I can say is that following Jason's advice (and a few others, use the board's search function), I was able to get rid of the micro-stutters I've had before.

One of Jason's advices was to use the "High Performance" Power Plan setting.

Since this isn't usually what you want to run all the time, you have the choice to either switch Power Plans manually whenever you run IL-2, or use a tool that switches the Power Plan automatically for you. That's one of the things FUMS can do.

Your statement was that Windows should do so automatically and that it would in some way compare to Intel's Turbo Boost.

Both is bull.

[ZachariasX]

Thank you. I didn't mean to make you angry by taking a dump on such tools, sorry if I have. And I am grateful that you do contribute useful things to  the community. And I very much acknowledge the use for people like you who work with the "inner things" of a computer.

 

But I do think the OP make a clear tone that this

42 minutes ago, SAS_Storebror said:

Did anyone claim such?

is the general expectation.

 

Most just install the tool and hope for more FPS. I don't think your tool does that. But there are plenty of tools pretending to do so while being far less honest about what it does than you do.

 

Regarding

48 minutes ago, SAS_Storebror said:

One of Jason's advices was to use the "High Performance" Power Plan setting. 

this is of course a sound advice, but you will agree that this is not really due to CPU scaling issues, but much rather system drivers that create lag hicups when sending peripheral diveses to sleep or waking them up. "Balanced" or "performance", CPU clocks remain the same.

 

Now in the light of jasons advise, I would say most consistent performance is obrtained by disabling any turbo and standby modes in BIOS. Prepar3D is a microstutterfest if you are running excessive settings. Just nailing CPU frequency on all cores helps a great deal in getting smooth performance. But on the other hand, it gives you a computer that you don't want to do things other than running a flight sim. Saving BIOS presets helps. But I guess this also is a non-advise as one shouldn't encourage people (that don't do so on their own) messing with the BIOS. So yes, click for "performance" and hope that helps.

 

So, no offense intended, you presented a tool in the context of "performance enhanercs" and I bit on that.

 

[JonRedcorn]

Switching to high performance mode can alleviate the computer causing hitching from switching back and forth from a low clock state to full clock state. This can induce hitching. Using that tool you could run the computer in the save energy mode and have it flip on to high performance when the game comes on, not sure whats so hard to understand here, this isn't a memory enhancer tool.

[SAS_Storebror]

1 hour ago, ZachariasX said:

I didn't mean to make you angry by taking a dump on such tools, sorry if I have.

 

Alright, let's be friends again

FUMS is not supposed to give higher FPS, that's right.

It has started years ago as a "process and affinity settings tool" privately built in order to run an IL-2 1946 dedicated server on a shared/virtual host.

At that time, it was essential to set the affinity of all CPU intensive processes (IL-2 1946 Server executable, server manager, webserver, database, others) so they don't step on each others' toes. And it helped a lot to raise the priority for the server executable so that in doubt, the other processes would have to wait but the server's main thread would keep running uninterrupted.

 

IL-2 Great Battles is different and so are the requirements to operate a server and, in other dimensions but to a similar effect, the game client. That's where the tool got reworked again.

The main feature of the rework from "prio/affi tool" to FUMS was to add the automatic Power Plan selection based on processes starting/running/terminating.

This was to follow Jason's advise to run your PC in "High Performance" Power Plan when you run the game (recommended) or dserver (absolutely essential here).

The other features (core affinity, process priority) don't seem to have any noticeable effect for IL-2 Great Battles. They're not mentioned by Jason either so we are in accordance with each other.

 

Changing your Power Plan can have a massive impact on the game's smoothness.

Technically, several Power Plan settings are important here, like wireless powersavings (if your PC is connected to Wifi and you play online), selective USB powersaving (can have all kind of impact) but most notably PCI Express energy management (enable/disable L0S or L1 modes, can cause massive GPU lag), minimum CPU power (5% in balanced mode, changing clock rates can cause lag and therefore micro stutters) and cooling rules.

 

All of this gets handled by FUMS automatically, i.e. when you crank up IL-2.exe, your system can automatically switch to "High Performance" or even "Ultimate Performance" mode, and it will automatically switch back to your default mode (usually "Balanced") when you end the game.

 

Mike

[JG_deserteagle540]

Hi Mike,

 

 

This FUMS tool works great.

 

Thank you very much.