FTB performance optimization

  • Please make sure you are posting in the correct place. Server ads go here and modpack bugs go here
  • The FTB Forum is now read-only, and is here as an archive. To participate in our community discussions, please join our Discord! https://ftb.team/discord

Bigglesworth

New Member
Jul 29, 2019
1,072
0
1
Not entirely true. JVM arguments CAN improve your gameplay experiences and sometimes even FPS. Java running with default settings isn't optimized for specific Apps/systems at all. So adjusting the JVM can improve your overall performance a bit.
But one should not expect increases in performance like when you update from an i3 to an i7 that is true. MC is for the most part crippled by the CPU but you can help your poor CPU a bit.


I've found no parameters I can change to improve framerate. Usually if you have parameters wrong, you simply either have your game crash, or performance is abysmal as the game starts paging instead of using RAM. Ive never encountered a case where Im getting, say, 40FPS, and I tweak a parameter and get 45. Maybe there is a "Use moar threads" param Im missing? :p

If people want to realistically improve framerate without spending $, they need to OC. If they already OC and have a 4+ core system, they need to find a utility that allows them to OC 2 cores and let the other cores stay at stock. You can OC those two cores higher. Then you go into taskman and set java.exe affinity to those two cores so windows doesnt pass the threads around to slower cores. Boom 10%-25% higher framerates

[DOUBLEPOST=1371586274][/DOUBLEPOST]
Ordered it online? I hate ordering stuff online but its just more convenient than driving around store to store to find parts, i remember when i first built this maybe 2 years ago from scratch, ordered everything online from different websites and it was a real ball ache.

Why would anyone hate ordering stuff online?
1. Cheaper
2. Huge selection to choose from and competition between online stores
3. Massive amounts of reviews so you know the real quality of what youre getting
4. Delivered to your door

Unless you have a microcenter, or cant wait 2 days, there is no reason to not go online for parts.
 

FMAylward

Active Member
Jul 29, 2019
68
0
26
Just going to say, with my system in a fresh world when there none/very little machines optfine will actually lower my FPS but once I have a base setup it took me from 10-15 to constant 30 fps. I am guessing it would have been higher if I was not limiting it to 30.
 

Ember Quill

Well-Known Member
Nov 2, 2012
350
119
68
I've found no parameters I can change to improve framerate. Usually if you have parameters wrong, you simply either have your game crash, or performance is abysmal as the game starts paging instead of using RAM. Ive never encountered a case where Im getting, say, 40FPS, and I tweak a parameter and get 45. Maybe there is a "Use moar threads" param Im missing? :p
If you've got RAM to spare, increasing the amount of RAM Minecraft can use tends to help A LOT. Not with framerate, exactly, but with most other sources of lag.

Optifine actually has a couple options that are essentially "use moar threads," assuming you use Optifine. And there are some JVM arguments that do something similar, but I don't know whether it actually affects performance, and if your specs aren't up to it, it might just make performance issues worse.
 

FMAylward

Active Member
Jul 29, 2019
68
0
26
Alloting too much RAM to MC will degrade performance. If you found a sweetspot, dont fix it. ;)

That might be because its being put into the pagefile/virtual memory instead of actual ram, a hard disk is a lot slower then ram. Any idea if anyone has done some kind of test for that?
 

Hydra

New Member
Jul 29, 2019
1,869
0
0
It is said that eye does 20 fps. .


It has never been said that the eye "does" 20 FPS and this is a myth that's as old as computergames itself.

Our brain starts accepting sequential still pictures as a continuous moving image starting at around 25 FPS. Lower than that and our brain just sees individual pictures. However; this doesn't mean that our eyes/brain can't see the difference between 24 and 60 FPS (or 60 and 120 actually). Especially in computer games when you turn around quickly there is a HUGE difference between say 30 and 60 FPS.

24 FPS is the minimum for our brain to not see a game as stuttering individual pictures. But you actually need a lot more in games for smooth gameplay.
 

FMAylward

Active Member
Jul 29, 2019
68
0
26
It has never been said that the eye "does" 20 FPS and this is a myth that's as old as computergames itself.

Our brain starts accepting sequential still pictures as a continuous moving image starting at around 25 FPS. Lower than that and our brain just sees individual pictures. However; this doesn't mean that our eyes/brain can't see the difference between 24 and 60 FPS (or 60 and 120 actually). Especially in computer games when you turn around quickly there is a HUGE difference between say 30 and 60 FPS.

24 FPS is the minimum for our brain to not see a game as stuttering individual pictures. But you actually need a lot more in games for smooth gameplay.


From what I can remember my game graphics university lecturer said about this basically agrees with all but one point there, his points were. We don't see in fps, gave no lower limit of "seeing in fps" and when I asked about people claiming to tell the difference ha basically said under 60fps believe them until proven wrong but if its over 60fps call them a liar until they prove otherwise.

I have to agree with his no lower limit as I can play minecraft at 15fps and it seems fine to me but only if its a constant 15fps, if its jumping about then it starts to feel off/wrong to me.
 

Ember Quill

Well-Known Member
Nov 2, 2012
350
119
68
That might be because its being put into the pagefile/virtual memory instead of actual ram, a hard disk is a lot slower then ram. Any idea if anyone has done some kind of test for that?
Paging DEFINITELY slows things down, but not because a hard disk is slower. Rather, it's because it's swapping data between the RAM and the HDD as required, which is an extra operation that slows things down. Depending on your HDD speed, fragmentation, latency, and loads of other factors, paging can become a serious bottleneck.
 

Poppycocks

New Member
Jul 29, 2019
1,914
0
0
It has never been said that the eye "does" 20 FPS and this is a myth that's as old as computergames itself.

Our brain starts accepting sequential still pictures as a continuous moving image starting at around 25 FPS. Lower than that and our brain just sees individual pictures. However; this doesn't mean that our eyes/brain can't see the difference between 24 and 60 FPS (or 60 and 120 actually). Especially in computer games when you turn around quickly there is a HUGE difference between say 30 and 60 FPS.

24 FPS is the minimum for our brain to not see a game as stuttering individual pictures. But you actually need a lot more in games for smooth gameplay.
Does your reading comprehension really only apply till you read the first disputable point?
 

Hydra

New Member
Jul 29, 2019
1,869
0
0
Does your reading comprehension really only apply till you read the first disputable point?


I'm just explaining something about something that a lot of people get wrong? What's so bad about that? Do you really need to pull out an ad hominem just because I added on to something you said?
 

Bigglesworth

New Member
Jul 29, 2019
1,072
0
1
It has never been said that the eye "does" 20 FPS and this is a myth that's as old as computergames itself.

Our brain starts accepting sequential still pictures as a continuous moving image starting at around 25 FPS. Lower than that and our brain just sees individual pictures. However; this doesn't mean that our eyes/brain can't see the difference between 24 and 60 FPS (or 60 and 120 actually). Especially in computer games when you turn around quickly there is a HUGE difference between say 30 and 60 FPS.

24 FPS is the minimum for our brain to not see a game as stuttering individual pictures. But you actually need a lot more in games for smooth gameplay.

It doesn't work quite like that. 24FPS was not chosen for anything about human anatomy or human minimums or maximums. It was chosen for far more practical reasons having to do with ease of editing and divisibility and film manufacturing standards. It is still that number mainly to do with cost, and the fact that the entire industry is used to the number and has in large part evolved around it. Fact of the matter is it is simply 'good enough' for human vision when coupled with what cameras do naturally; blur movement. 12 FPS can seem just as smooth as 24 if slightly more motion blur is applied.

You will see as the technology for digital imaging and video display advances TVs will start making 60-120Hz the standards along with 4K HD. People will hate it, and rightly so, because a lot of the first productions that use the 120Hz abilities will forget something important: the shutter speed of what is being filmed needs to be adjusted so motion blur is reduced so production doesnt result in un-naturally smooth video. The whole point of higher framerate is to allow sharper images to successively produce the illusion of motion.

After 120 is drops off sharply and there isnt much point. The human eye can physically 'see' several hundred FPS, but there is a point where the receptive cells simply cannot react to changing light fast enough and will simply start to 'skip frames'.


Paging DEFINITELY slows things down, but not because a hard disk is slower. Rather, it's because it's swapping data between the RAM and the HDD as required, which is an extra operation that slows things down. Depending on your HDD speed, fragmentation, latency, and loads of other factors, paging can become a serious bottleneck.

... no. Its because the hard drive is much slower. The 'extra operation' is negligible. When information is needed from the hard drive, the operation itself to fetch is takes very little time (nanoseconds), the time to seek, and then to read/write it is what takes several orders of magnitude longer than if it were in RAM. If the hard drive were as fast as the RAM, you would see no slowdown, proving this point. The game will basically freeze several times a second while this happends as it cant continue without the needed information, resulting in 1-2 fps. Some games handle this more elegantly than others.... (like having placeholders in RAM at all times [very low-fi textures or just colors] so even if a needed texture is being swapped for some god-awful reason, the game can still at least render using a placeholder)
 

Hydra

New Member
Jul 29, 2019
1,869
0
0
It doesn't work quite like that. 24FPS was not chosen for anything about human anatomy or human minimums or maximums. It was chosen for far more practical reasons having to do with ease of editing and divisibility and film manufacturing standards.

I know, I work in broadcast ;) That's why I said 'about', it depends on a lot of circumstances. How easy it is to 'fool' your brain even depends on if you had your cup of coffee this morning :)
 

Poppycocks

New Member
Jul 29, 2019
1,914
0
0
I'm just explaining something about something that a lot of people get wrong? What's so bad about that? Do you really need to pull out an ad hominem just because I added on to something you said?
Can't be ad hominem if we agree with each other, which "adding on top of" implies. But yeah, sure, I guess.
 

Ember Quill

Well-Known Member
Nov 2, 2012
350
119
68
... no. Its because the hard drive is much slower. The 'extra operation' is negligible. When information is needed from the hard drive, the operation itself to fetch is takes very little time (nanoseconds), the time to seek, and then to read/write it is what takes several orders of magnitude longer than if it were in RAM.
Milliseconds, actually, not nanoseconds. There's a million nanoseconds in a millisecond, and most HDDs have a seek time ranging from 10-15 milliseconds. It doesn't seem like much, but it's significant. Latency adds a few more milliseconds to that.

But yeah, you're right about HDD speeds being drastically slower as well. I dunno how my math was so far off.

EDIT: Messed up the MT-to-Mb conversion ratio. That's how I got it wrong. Whoops.
 

Hydra

New Member
Jul 29, 2019
1,869
0
0
EDIT: Messed up the MT-to-Mb conversion ratio. That's how I got it wrong. Whoops.


Even if you didn't divide by 8 HDs are still an incredible amount slower than memory. The fastest drives now are around 90MB/sec (talking about harddrives, not SSDs ofcourse). Memory has a bandwidth of around 25GB per second these days.
 

Ember Quill

Well-Known Member
Nov 2, 2012
350
119
68
Even if you didn't divide by 8 HDs are still an incredible amount slower than memory. The fastest drives now are around 90MB/sec (talking about harddrives, not SSDs ofcourse). Memory has a bandwidth of around 25GB per second these days.
It's actually half that, around 12.5GB/s. Unless you're using Dual Channel.

I blame sleep deprivation for my rather massive mathematical errors. I feel quite stupid now.
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Even if you didn't divide by 8 HDs are still an incredible amount slower than memory. The fastest crummy notebook drives now are around 90MB/sec and desktop drives around 150MB/sec (talking about harddrives, not SSDs ofcourse). Memory has a bandwidth of around 25GB per second these days.

Fixed that for you ;)

Not that it changes anything about the point, though. Just indulging my German OCD.
 

Hydra

New Member
Jul 29, 2019
1,869
0
0
Fixed that for you ;)

Not that it changes anything about the point, though. Just indulging my German OCD.


Yeah, I've not been keeping up to date with current specs really. I only do that about every 4 years when my PC needs an update :)