USB polling precision - Page 118 - Overclock.net - An Overclocking Community

Forum Jump: 

USB polling precision

Reply
 
Thread Tools
post #1171 of 1186 (permalink) Old 07-31-2020, 01:01 AM
New to Overclock.net
 
empl's Avatar
 
Join Date: May 2016
Posts: 93
Rep: 1 (Unique: 1)
Quote: Originally Posted by Timecard View Post
Isn't DPC the software layer abstraction for queued interrupt requests?
And what with it? All i know of it: that cpu gets interrupts from hardware and than schedules DPC call. But higher the latency, than the longer time for processor to handle interrupts. You get frame each 16.67, or 6.9 ms etc. But what if interrupt is handled right after a frame is finished? Than it doesn't render last mouse position in that frame and you'll have to wait for a next frame to update another mouse position. If you have 500hz polling e.g. Than if mouse captures position each 2 ms. E.g. 2ms before gpu finishes rendering a frame (i doubt mouse position can be updated once frame is sent to monitor from gpu) and than on top of that: if you have high dpc. latency - it can take like another 100 us on good motherboards to handle interrupt from usb. Even usb doesn't send interrupts, but cpu polls it 500/1000 times per second. CPU still schedules dpc calls for USB driver... But than, if cpu handles interrupt 100us: after a frame is finished rendering and is sent to monitor. Than gpu had rendered mouse position from time 4ms ago, because new mouse position wasn't updated yet!!! Cpu prepares each frame for GPU, before GPU starts to render it, i don't know if cpu can update mouse position to gpu, once gpu has a frame. But after gpu sends it to monitor, it is doubtful. Than you can see over 1 second of time, mouse position can lag each time up to 144 times, if you use 144hz monitor. And small input lag adds up!

It doesn't affect you, that much, because you have 200 ms reaction time on average and if on top of overall input lag. 2.6 ms is added, it won't be major a difference. But you can feel it on overall mouse movement and especially: when you do small adjustments and mouse movement is clealer! Also movement becomes more stable, than if you have high dpc. latency. It is called micro stuttering!

That's why if you set timer resolution to 0.5ms, there is gigantic difference between 1ms. It allows to update code to cpu faster. Yes it is only 500us difference, but than it means: it can update code to cpu each 500us, which scales over 1 second of time, because you have 144fps e.g. Than if you could send code to cpu only each 1ms, or at higher interval... You can try it yourself. There is gigantic difference! Also if you will be using 500hz, instead 1000hz. If mouse position renders from previous frame (4ms ago), there will be larger feeling that mouse lags. That if it used latest position from (2ms ago). So even if you have 100 us dpc. latency and there is 144 fps. So there is chance 144 times: mouse cursor can be rendered not from a last position! Depending how polling occurs and dpc. latency, before interrupt is handled by cpu.

I am not hardware expert, but if there wasn't any difference, than i couldn't feel anything. You can try timer resolution 500us vs 1ms, there is so gigantic difference, that anyone can probably tell. A lot of people use it. Also you can try 15.6ms, vs 0.5us and you have to disable dynamic tick, so timer resolution is constant! Also it increases cpu load slightly. It was tested in Crysis to add fps.

Also about stability: if you use 0.5ms it feels more unstable, than 1ms. It is hard to describe, but 0.5 ms is more inconsistent than 1ms, especially with 500hz polling. With 1000hz polling, it will be better. That's reason i was always using 1ms timer resolution, feels best! And 500hz also better than 1000hz. Because 1000hz captures even smallest movements... So 500hz is more accurate, but less consistent!

DPC lat. matters, Anandtech mobo DPC lat. tests, My input lag guide - https://www.tenforums.com/gaming/117...7=#post1454596

Last edited by empl; 07-31-2020 at 01:18 AM.
empl is offline  
Sponsored Links
Advertisement
 
post #1172 of 1186 (permalink) Old 07-31-2020, 01:13 AM
New to Overclock.net
 
x7007's Avatar
 
Join Date: Aug 2009
Posts: 1,208
Rep: 12 (Unique: 12)
Quote: Originally Posted by empl View Post
Quote: Originally Posted by Timecard View Post
Isn't DPC the software layer abstraction for queued interrupt requests?
And what with it? All i know of it: that cpu gets interrupts from hardware and than schedules DPC call. But higher the latency, than the longer time for processor to handle interrupts. You get frame each 16.67, or 6.9 ms etc. But what if interrupt is handled right after a frame is finished? Than it doesn't render last mouse position in that frame and you'll have to wait for a next frame to update another mouse position. If you have 500hz polling e.g. Than if mouse captures position each 2 ms. E.g. 2ms before gpu finishes rendering a frame (i doubt mouse position can be updated once frame is sent to monitor from gpu) and than on top of that: if you have high dpc. latency - it can take like another 100 us on good motherboards to handle interrupt from usb. Even usb doesn't send interrupts, but cpu polls it 500/1000 times per second. CPU still schedules dpc calls for USB driver... But than, if cpu handles interrupt 100us: after a frame is finished rendering and is sent to monitor. Than gpu had rendered mouse position from time 4ms ago, because new mouse position wasn't updated. Cpu prepares each frame for GPU, before GPU starts to render it, i don't know if cpu can update mouse position to gpu, once gpu has a frame. But after gpu sends it to monitor, it is doubtful.

That's why if you set timer resolution to 0.5ms, there is gigantic difference between 1ms. It allows to update code to cpu faster. Yes it is only 500us difference, but than it means: it can update code to cpu each 500us, which scales over 1 second of time, because you have 144fps e.g. Than if you could send code to cpu only each 1ms, or at higher interval... You can try it yourself. There is gigantic difference! Also if you will be using 500hz, instead 1000hz. If mouse position renders from previous frame (4ms ago), there will be larger feeling that mouse lags. That if it used latest position from (2ms ago). So even if you have 100 us dpc latency and there is 144 fps. So there is chance 144 times: mouse cursor can be rendered not from a last position! Depending how polling occurs and dpc. latency, before interrupt is handled by cpu.

I am not hardware expert, but if there wasn't any difference, than i couldn't feel anything. You can try timer resolution 500us vs 1ms, there is so gigantic difference, that anyone can probably tell. A lot of people use it. Also you can try 15.6ms, vs 0.5us and you have to disable dynamic tick, so timer resolution is constant! Also it increases cpu load slightly. It was tested in Crysis to add fps.

Also about stability: if you use 0.5ms it feels more unstable, than 1ms. It is hard to describe, but 0.5 ms is more inconsistent than 1ms, especially with 500hz polling. That's reason i was always using 1ms timer resolution, feels best! And 500hz also better than 1000hz. Because 1000hz captures even smallest movements... So 500hz is more consistent and accurate!
disable dynamic tick yes is bad. it changes the mouse accelerating or decelerating. I tested it and someone else in Tom'shardware. I told him and he noticed it. shouldn't be used.

RIGWarning: Spoiler! (Click to show)
CPU :AMD 1950X @ 3.7 Ghz 1.280
MOBO : Gigabyte X399 AORUS Gaming 7 (rev. 1.0)
RAM : G.SKILL Flare X Series 32GBx4 DDR4 3200 (PC4 25600) Model F4-3200C14D-16GFX)
GPU :INNO3D Nvidia GeForce GTX1080 8G D5 1759Mhz iChill X3
Sound Card For 5.1 Speakers : Asus Strix RAID DLX
Network Card : SolareFlare SFN6122F SFP+ 10Gbit Dual Port PCIEx4
PSU : ThermalTake 850Watt ToughPower
STORAGE
SanDisk Extreme PRO SSD 480Gb OS
WD Gold 6TB DataCenter 7200 128MB DATA/GAMEZ
Seagate 5TB Enterprise 7200 128MB (ST5000NM0024) DATA/GAMEZ
Samsung x2 1TB 7200 RAID0 GAMEZ
Segate 3TB Backup+ Desktop USB External BACKUP

GEAR and Sound
Warning: Spoiler! (Click to show)
TV : LG OLED55C6 TV
Mouse : Logitech G502
Mouse Pad : SteelSeries I-2 Black
Keyboard : Corsair Gaming K70 RAPIDFIRE
GamePad : Xbox Elite Controller + Xbox Wireless Adapter
Headphones : Beyerdynamic DT990 600 Ohms
A...


x7007 is offline  
post #1173 of 1186 (permalink) Old 07-31-2020, 01:22 AM
New to Overclock.net
 
empl's Avatar
 
Join Date: May 2016
Posts: 93
Rep: 1 (Unique: 1)
Quote: Originally Posted by x7007 View Post
disable dynamic tick yes is bad. it changes the mouse accelerating or decelerating. I tested it and someone else in Tom'shardware. I told him and he noticed it. shouldn't be used.
Yeah i can't wait btw: until some random peoples start jumping in and telling i am crazy and i am imagining it... Which probably think: that eye can't see more than 24 fps. Scientists discovered eye can see even one photon of light. Why it is so hard for people to believe, that people can tell input lag? And why the heck, someone cares, if someone else thinks otherwise???! If you think it doesn't help, why do you care and you have to flame other people, if they have different opinion?

Btw you may take interest in this:
https://csgotweaks.com/bios.html

Also did you try to delete wan adapters in device manager and do you disable usb power saving there, on top of in power plan setting? I don't know why, even device is not used, some drivers introduce input lag. Like if you use virtual mechanic, but you have nothing mounted. Or even i don't use Nvidia HD audio, disabling it reduced input lag! I also though of lowering priority of some useless system services, in process lasso. Which cannot be safely disabled. Or reducing core affinity. Because they sometimes consume cpu cycles. And cpu scheduling is pretty good. I still play competitive games sometimes and i have VR, so i enjoy low input lag. Btw Valve Index Hardware has very low input lag, i was surprised!

This may amuse you: https://forums.blurbusters.com/viewt...p=55454#p55454 Also monitor expert telling color profiles can reduce input lag, or using no color profile. After someone flaming, that it doesn't matter haHAHHAHA...

DPC lat. matters, Anandtech mobo DPC lat. tests, My input lag guide - https://www.tenforums.com/gaming/117...7=#post1454596

Last edited by empl; 07-31-2020 at 01:30 AM.
empl is offline  
Sponsored Links
Advertisement
 
post #1174 of 1186 (permalink) Old 07-31-2020, 02:27 AM
New to Overclock.net
 
numberfive's Avatar
 
Join Date: Aug 2016
Posts: 625
Rep: 22 (Unique: 12)
Quote: Originally Posted by empl View Post
How is that zero difference, you have these motherboards - even 500$ asus rogs, which have literally 1ms dpc latency. If motherboard is hogged with dpc latency, other devices suffer too. Besides on cheap motherboard you may not be able to disable individual USB ports. I found out on my ASUS Z390-gaming i, when i disable all usb ports except usb 2: i have USB Root Hubs. But if i enable even 1x usb 3.0: i have USB Root Hubs (USB 3.0) and mouse movement feels like crap (while i am using USB 2 ports for mouse and keyboard)! Also if you aren't able to set bclk closest to 100, it is not optimal. Or if you are unable to disable HPET, which is not possible on asus motherboard - not sure if this is true for all... Also i don't know if all modern moderboards use msi, which is huge deal for usb polling stability, or latency. So there will be difference for sure. There will be more difference in input lag probably (depending what features your motherboard can tweak) and dpc latency, which depends on what mobo you choose and what devices/drivers it is using...

I don't think polling stability is even main issue, unless you suffer from high dpc. latency! But i think input lag is more important! Today if you choose decent motherboard and disable useless services etc: polling stability will be probably very consistent. Bigger problem is choosing right mobo with low dpc. latency and features you can tweak. Majority of input lag comes from untweaked bios. What concerns stability not sure, i don't test BIOS features for stability, i already know what should be disabled etc. And mine is very decent. Not sure if that's a reason, but on my new motherboard supporting msi for USB ports - mouse polling stability is much better and there is a lot lower dpc. latency for USB driver, than on my old one! By margin like 100 us, depends.

ANANDTECH HAS A LOT OF TESTS FOR MOBOS INCLUDING DPC LATENCY! It makes all difference in the world!!! I was supreme master class in cs go and when i tried USB 3, i couldn't hit anything, i would be bronze Some of these settings have gigantic impact on mouse, like disabling CSM helps a ton!!!
none of this matters. price of the motherboard has nothing to do with the quality of the hardware and firmware, many mobo components (like usb controllers) would be the same on budget and luxury boards, it's not unusual for an expensive mobo to have a ****ty component embedded for no reason as well.
I mean you need to have at least a hint of justification for the dpc latency tantrum. I don't see it, never seen a single test where 10us was practically better than 100us for gaming, not a single one.

Quote: Originally Posted by empl View Post
Btw you may take interest in this:
https://csgotweaks.com/bios.html
garbage.

like from the beginning "100:100. 100:100 should obviously provide better results since it matches BCLK. Memory with lower frequency and lower timing can reduce latency."
what is obvious about it? if number is the same it doesn't matter that they are in sync, and if they would be in sync, it doesn't make it better for some reason.
"The closest you can get this to 100.00, the better." no? nothing better or worse about this.
etc

Roach's declowning guide to the galaxy https://i.imgur.com/iz6WDUl.jpg

Last edited by numberfive; 07-31-2020 at 03:51 PM.
numberfive is offline  
post #1175 of 1186 (permalink) Old 07-31-2020, 02:51 AM
New to Overclock.net
 
Join Date: Nov 2009
Location: µcode
Posts: 819
Rep: 60 (Unique: 49)
Quote: Originally Posted by x7007 View Post
disable dynamic tick yes is bad. it changes the mouse accelerating or decelerating. I tested it and someone else in Tom'shardware. I told him and he noticed it. shouldn't be used.
Last time I looked at this if you disable idle states then dynamic tick is disabled too as there is no opportunity to turn it off however IIRC disabling it by BCD see's the tick on only core 0 (on my HW anyways) which could cause conflicts of interest.

Timer resolution affects quantum's (thread time) so making it very low may affect performance in some situations depending on software etc then again being low means W10 will be messing about with it a lot less hopefully.

As for the more expensive mainboards, they might provide better overclocking and the faster your CPU runs the quicker it can get things done, generally speaking.

ucode is offline  
post #1176 of 1186 (permalink) Old 07-31-2020, 03:16 AM
New to Overclock.net
 
empl's Avatar
 
Join Date: May 2016
Posts: 93
Rep: 1 (Unique: 1)
Quote:
garbage.

like from the beginning "100:100. 100:100 should obviously provide better results since it matches BCLK. Memory with lower frequency and lower timing can reduce latency."
what is obvious about it? if number is the same it doesn't matter that they are in sync, and if they would be in sync, it doesn't make it better for some reason.
"The closest you can get this to 100.00, the better." no? nothing better or worse about this.
etc
I never said it has. I was listing it as example. That even rogs for 500$ suffer from high dpc. latency (even 1ms). I mentioned this... It doesn't have to do with quality of hardware, but rather from poorly written drivers! Hardware plays role too, e.g. motherboard rgb was found to cause latency, or wifi. So you need hardware/bios, which can allow you to tweak and disable options. Someone claims, even VRM switching matters for input lag. But for dpc. latency solely, it is drivers which cause delay! So picking right hardware (motherboard - makes all difference in the world)! Again you don't even read and just berate and flame... I am not sure, if all new motherboards support msi for USB and if you can turn off individual USB ports on them. You are just cherry picking facts to flame more... If you don't care about it, why you care some people can tell difference ? Even if you had 10us and 100us. Than sometimes: latest mouse position won't render. And than you will experience less stability and cursor smoothness. Even in this case, it wouldn't probably matter much. But if it goes to extremes like 1ms of dpc. latency, than you will have micro-stuttering. And i also said, if you read... That it won't probably matter as much as input lag (depends on hardware/bios again, so what hardware you pick matters), if you have already decent motherboard, which has 100 us dpc. latency for USB driver. So you just don't read and than claim, i said things - i didn't say. So you just troll... And yes if you play on high level competitive games, that every distraction matters and it helps to have less input lag etc. Previously i was gold 3 in cs go and after tweaking my system, i instantly ranked up. And i tried previous setting and i couldn't hit anything...

IT IS MEANINGLESS TO TALK WITH YOU, YOU JUST FLAME, INSTEAD OF DISCUSSION... I am just calling you out...

DPC lat. matters, Anandtech mobo DPC lat. tests, My input lag guide - https://www.tenforums.com/gaming/117...7=#post1454596

Last edited by empl; 07-31-2020 at 03:27 AM.
empl is offline  
post #1177 of 1186 (permalink) Old 07-31-2020, 06:32 AM
New to Overclock.net
 
Timecard's Avatar
 
Join Date: Dec 2015
Posts: 499
Rep: 32 (Unique: 20)
Quote: Originally Posted by empl View Post
It doesn't have to do with quality of hardware, but rather from poorly written drivers!
This was my point about DPCs and you saying buy motherboards with low dpcs, just wanted to make sure readers understood. Sooo.... I think what you were trying to say is some motherboards operate better with the available drivers assuming Anandtechs (your reference) tests aren't unknowingly skewed.

numberfive is just stating that many components operate on the same clock so as long as the clock is stable they are equally synchronized, seems logical. I assume you'll get some performance benefit with having it higher e.g. constant 99.9 vs 100 but it shouldn't be significant. If anyone has evidence or research for the contrary please post.

Last edited by Timecard; 07-31-2020 at 06:46 AM.
Timecard is online now  
post #1178 of 1186 (permalink) Old 07-31-2020, 07:18 AM
New to Overclock.net
 
numberfive's Avatar
 
Join Date: Aug 2016
Posts: 625
Rep: 22 (Unique: 12)
Quote: Originally Posted by empl View Post
I never said it has. I was listing it as example. That even rogs for 500$ suffer from high dpc. latency (even 1ms). I mentioned this... It doesn't have to do with quality of hardware, but rather from poorly written drivers! Hardware plays role too, e.g. motherboard rgb was found to cause latency, or wifi. So you need hardware/bios, which can allow you to tweak and disable options. Someone claims, even VRM switching matters for input lag. But for dpc. latency solely, it is drivers which cause delay! So picking right hardware (motherboard - makes all difference in the world)! Again you don't even read and just berate and flame... I am not sure, if all new motherboards support msi for USB and if you can turn off individual USB ports on them. You are just cherry picking facts to flame more... If you don't care about it, why you care some people can tell difference ? Even if you had 10us and 100us. Than sometimes: latest mouse position won't render. And than you will experience less stability and cursor smoothness. Even in this case, it wouldn't probably matter much. But if it goes to extremes like 1ms of dpc. latency, than you will have micro-stuttering. And i also said, if you read... That it won't probably matter as much as input lag (depends on hardware/bios again, so what hardware you pick matters), if you have already decent motherboard, which has 100 us dpc. latency for USB driver. So you just don't read and than claim, i said things - i didn't say. So you just troll... And yes if you play on high level competitive games, that every distraction matters and it helps to have less input lag etc. Previously i was gold 3 in cs go and after tweaking my system, i instantly ranked up. And i tried previous setting and i couldn't hit anything...

IT IS MEANINGLESS TO TALK WITH YOU, YOU JUST FLAME, INSTEAD OF DISCUSSION... I am just calling you out...
- they are not suffering from high dpc latency, because the only people who care about high dpc latency are those who play 3dmark 2040 and MAYBE sound engineers in some exotic cases in 2006.
- quality of drivers has nothing to do with the cost of the hardware (which for some reason was your point two comments above)
- "Someone claims, even..." someone is even roach. ignore and forget, or prove and shut me up.
- "motherboard - makes all difference in the world" no? there is no right one, there are very few that are garbage and there are the rest, that will suit you perfectly for gaming.
- "Again you don't even read and just berate and flame..." for some reason you are spreading this bs and giving a link to a tweaking resource that is referring back to roach's thread on OCN. like for real? how can you not be a clown here?
- "If you don't care about it, why you care some people can tell difference ?" don't make me laugh. you are continuously clowning here as an assistance manager to roach.

I actually decided to not read your post further, because you repeat your bs claims which have zero justification, zero practical impact on (competitive) gaming and zero tests that would show even a slightest difference.
it is not a matter of what kind of difference I can tell or not, it is not about me at all.
Y O U
A R E
S P R E A D I N G
B S


on a side note, you are simply bad in games (according to you), you can improve with more practice and less tin foil tuning. stock windows 10 setup is not providing some disadvantageous experience to professional players, if it does to you, well, it is a you problem.

Roach's declowning guide to the galaxy https://i.imgur.com/iz6WDUl.jpg
numberfive is offline  
post #1179 of 1186 (permalink) Old 07-31-2020, 07:06 PM
New to Overclock.net
 
Join Date: Nov 2009
Location: µcode
Posts: 819
Rep: 60 (Unique: 49)
Quote: Originally Posted by Timecard View Post
numberfive is just stating that many components operate on the same clock so as long as the clock is stable they are equally synchronized, seems logical.
Looks at own board and sees lots of clocks that are not the same/synchronized, which ones did you mean? As for price not affecting firmware, that in my experience is just not true. A bit like Windows maybe, you pay more to use more features such as amount of RAM, it's just OS software settings.

ucode is offline  
post #1180 of 1186 (permalink) Old 08-01-2020, 06:50 AM
New to Overclock.net
 
Join Date: Jan 2020
Posts: 5
Rep: 0
I've depleted my tweaks ideas, that's all i can get @ win 10 (not safe mode).
Can't go any further with ryzen...
Attached Thumbnails
Click image for larger version

Name:	G pro tweaked.png
Views:	21
Size:	61.7 KB
ID:	363420  

MIETAS is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off