Couldn't find out which USB ports are the CPUs & which are the Chipset's

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
phaze
Posts: 38
Joined: 23 Feb 2020, 20:37

Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by phaze » 14 Apr 2023, 08:54

Hey there,

I'm a pro VALORANT player and been into these things since way back, figuring out different things from time to time, came across the USB Ports could be causing slight input lag as it takes more time to transmit and receive when it's connected to the chipset's controller rather than the CPU's, my motherboard says

CPU:
3 x USB 3.2 Gen 1 ports on the back panel
1 x USB 3.2 Gen 2 Type-A port (red) on the back panel
Chipset:
1 x USB Type-C™ port on the back panel, with USB 3.2 Gen 2 support
1 x USB 3.2 Gen 2 Type-A port (red) on the back panel
2 x USB 3.2 Gen 1 ports available through the internal USB header
2 x USB 2.0/1.1 ports on the back panel
Chipset+2 USB 2.0 Hubs:
8 x USB 2.0/1.1 ports (4 ports on the back panel, 4 ports available through the internal USB headers)

there are many of them which are the same, so I had to use USB Device Tree Viewer to figure out, Now I want to know which CPU controller is which, the top one or the lower one, I assume its the top cuz it has the prioritized PCI Busing which is Number 1, while the other one is PCI Bus 7

also, I used Device Cleanup tool to clean all unused device entries.


Also, Higher DPI causes less input lag than 400 DPI for example? or is it the same since the mouse displacements are the same?
for example

Which is better?
400 DPI and 0.8 ingame OR 1600 DPI and 0.2 ingame?


EDIT: Came across a Topic that says I should use triple digits+ in ingame sensitivity, for example @400dpi - 0.800 or 0.801 if it doesn't work, did I understand this correctly? or I misunderstood something!

<3
Attachments
gafasdaf.png
gafasdaf.png (17.09 KiB) Viewed 9078 times


Boomchakadah
Posts: 197
Joined: 15 Aug 2018, 02:44

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Boomchakadah » 15 Apr 2023, 05:32

Which motherboard do you have?
VG258QM 280Hz TN (main)| XG2431 240Hz IPS | XL2540K 240Hz TN | XV252Q 1080p 280Hz IPS | XL2546K 1080p 240hz TN | AW2518HF 1080p 240Hz TN | XV240YP 1080p 165Hz IPS | XG2402 1080p 144hz TN | 27GL83A 1440p 144Hz IPS | XL2411P 144Hz TN | XF240H 144Hz TN

Calypto
Posts: 17
Joined: 16 Sep 2019, 20:58

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Calypto » 15 Apr 2023, 12:29

You can easily verify the port by going into hwinfo main window→bus→check which controller is which; with Zen it will show the CPU and chipset controllers separately.
Capture.PNG
Capture.PNG (21.01 KiB) Viewed 8657 times

espresso
Posts: 63
Joined: 11 Feb 2022, 04:56

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by espresso » 15 Apr 2023, 13:33

Calypto wrote:
15 Apr 2023, 12:29
You can easily verify the port by going into hwinfo main window→bus→check which controller is which; with Zen it will show the CPU and chipset controllers separately.
Capture.PNG
And then how would you know which port is which on the back of the MB?

Calypto
Posts: 17
Joined: 16 Sep 2019, 20:58

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Calypto » 15 Apr 2023, 15:14

Trial and error.

User avatar
Chief Blur Buster
Site Admin
Posts: 11708
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Chief Blur Buster » 15 Apr 2023, 18:02

phaze wrote:
14 Apr 2023, 08:54
Also, Higher DPI causes less input lag than 400 DPI for example? or is it the same since the mouse displacements are the same?
You need to reinterpret this differently:

When you move your mouse slowly, it may take longer for a mouse to register a movement if the DPI is low. So if you take 10 millisecond to move a mouse 1/400th of an inch (400dpi = 400 Dots Per Inch), it means you've got 10 milliseconds lag at 400dpi. Slow mouse movements = somewhat more laggy.

So raise to 800dpi or 1600dpi, and you have less lag for slow mouse movements.

Also, you get a higher mouselook framerate at a higher dpi. Today's monitors are hitting 500Hz+, and moving 0.5 inches per second at 400dpi gives you only 200 mouselook frames per second -- and very slow tracking-style mouselooks can be slower -- you know, how a sniper homes on a slow moving target by dragging mouse slower than a fast flick? -- bingo. Low dpi can also sabotage your slow-mouse-movement framerate. Grainy steppy-steppy slow tracks are more of a problem with low DPI settings.

And as you have slow movements (more lag due to longer time for the mouse to move 1/400sec, at least versus other higher DPI settings 1/800sec versus 1/1600sec) and fast movements, 400dpi has bigger mouse lag yo-yo effect than 800dpi, since mouse input during slow turns will less likely be lagged relative to during fast turns. So if you are at 1600dpi, your mouse only needs to move 1/1600 inch in order for a mouse report to arrive to the game engine -- and that can happen sooner.

But older sensors had a hard time at high DPI, so you need a recent mouse sensor, a good mousepad, and clean mouse feet, to get reliable high DPI.

Valorant is very friendly to high DPI, so give that a try -- 4x DPI at 1/4th sensitivity -- and your fast flicks will be unchanged feel. The dramatic improvements you will have gotten will be from your slowtracks (e.g. aiming a sniper scope on a slow-moving target). At today's refresh rates, even going above 1600dpi is now worth it if you can preserve your 400dpi-league muscle memory with proper mouse settings. You can use websites such as https://www.mouse-sensitivity.com/ to make sure your muscle memory is sync'd between different games, because all those different games and different sensitivities can wreak havoc with how your aimtraining memory is trained.

Usually, DPI will (usually) have little effect on fast aimtraining if you train using an aimtrainer, as long as your DPI and sensitivity is sufficiently high enough for the aimtrainer resolution / object sizes you're using. However DPI will have significant effect on other tactics such as slow tracking and more clearer slowturns (e.g. eye tracking the panning screen during slow turns), for the other types of game tactics that you may also do.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

User avatar
Chief Blur Buster
Site Admin
Posts: 11708
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Chief Blur Buster » 15 Apr 2023, 18:12

Calypto wrote:
15 Apr 2023, 15:14
Trial and error.
Yes, USB port roulette.

Here's the thread:
viewtopic.php?f=10&t=7618

However, whether the ports go through CPU or chipset or PCIe lanes is more academic to the fact that you have USB contention (processing/bandwidth-sharing) between the high-poll keyboard and high-poll mouse. That's the more important optimization you need to make -- preventing USB competition between USB devices is more important than caring about CPU vs chipset vs PCIe lanes.

You may even have slightly higher lag but predictable latency can be more important than erratic latency jitter -- low lag can becomes useless if there's a lot of latency variances (spikes, lag overshoots, lag undershoots, weird hitreg where hits don't register, or missed hits register). It's better to game at "higher-but-predictable" lag versus "low-but-erratic-hitreg" lag. More people quit esports from the latter problem.

It's best to have both, but if you had to pick your poison, now you know which poison to take...
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

User avatar
Chief Blur Buster
Site Admin
Posts: 11708
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Chief Blur Buster » 15 Apr 2023, 18:14

phaze wrote:
14 Apr 2023, 08:54
Which is better?
400 DPI and 0.8 ingame OR 1600 DPI and 0.2 ingame?
Triple digits is simply to fix your error margins.

For example some people use 400 dpi at 0.5 sensitivity.
The problem of going 1600 dpi requires 0.125 (preserves aimtraining memory) instead of 0.1 or 0.2 (wrecks your aimtraining memory)

If you are already 0.8 then the triple digit talk doesn't matter, it's irrelevant to you -- to you specifically in particular if you are using 400dpi at 0.8. But not everyone uses the same sensitivity settings as you do.

It's common sense math and science;

______

That said, not all engines may be friendly at all possible DPI settings at all possible sensitivity values.

Some people report weird behaviors in some game engines such as CS:GO at extremely high DPIs, but not everyone. While 400 and 800dpi are the classic standard for older games like CS:GO and 1600dpi did not work well in that game -- reportedly, it is possible to do 1600dpi accurately in CS:GO if you edit the configuration file to sync up your settings, and stick to mathematically clean numbers -- But CS:GO uses lower precision mouse math than Valorant, as one of the many culprits. This probably won't be an issue if your decimal digits count is the same at 400dpi as at 1600dpi, which minimizes opportunity for math rounding errors to build up in older, low-precision game engines. (...Inside a game engine at 500fps, your mouse math rounding errors may potentially build up faster than at 100fps, depending on how the game does its movement and physics...)

This is likely not going to be an issue if you use simple sensitivity numbers like you are (0.8 vs 0.2) but could become an issue if you're using weird numbers like 0.7 at 400dpi, and then you've got 0.175 for 1600dpi which may create more rounding errors at extremely high framerates (CS:GO already has a tendancy to behave wonky when run at 1000fps in bot tests, and some have mentioned that 1600dpi push the wonky framerate threshold lower).

It's currently easier to preserve memory in Valorant going from 400dpi to 1600dpi than it is when going from 400dpi to 1600dpi in CS:GO, as it is not as easy to optimize (especially if you're already using an odd divisor at 400dpi and/or you don't manually edit the configuration file, etc) -- there are more things that can go wrong/erratic/wonky at 1600dpi in CS:GO than in Valorant. Valorant is very easy/friendly to configure to 1600dpi and sync your muscle memories with other games.

The point being is, there are people reporting 400->1600dpi behaving wonkier in games using certain very old game engines like CS:GO than in Valorant, and there are people reporting no difference. At the end of the day, for the configuration-inexperienced person, going 1600dpi is pretty easy/safe in Valorant even just via the menus, but requires some mindfulness/testing/configuration file editing in CS:GO to make sure it doesn't feel different for fast flicks.

Be noted, some use numbers like 2000dpi or other numbers; as modern mouse sensors are more flexible to configure; although you have to be careful to preserve the mouse mathematics -- use a mouse sensitivity calculator and you should be OK.

For this "unpredictable glitchinesses" reasons -- that's is usually why if you look at certain websites such as prosettings.net, you will see nobody on that seems to use above 800dpi in old engines such as CS:GO but that it is more common in Valorant or Fortnite by the professional esports teams to use quadruple-digit DPI in the newer engines.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Slender
Posts: 632
Joined: 25 Jan 2020, 17:55

Re: Couldn't find out which USB ports are the CPUs & which are the Chipset's

Post by Slender » 15 Apr 2023, 21:39

phaze wrote:
14 Apr 2023, 08:54
Hey there,

I'm a pro VALORANT player and been into these things since way back, figuring out different things from time to time, came across the USB Ports could be causing slight input lag as it takes more time to transmit and receive when it's connected to the chipset's controller rather than the CPU's, my motherboard says

CPU:
3 x USB 3.2 Gen 1 ports on the back panel
1 x USB 3.2 Gen 2 Type-A port (red) on the back panel
Chipset:
1 x USB Type-C™ port on the back panel, with USB 3.2 Gen 2 support
1 x USB 3.2 Gen 2 Type-A port (red) on the back panel
2 x USB 3.2 Gen 1 ports available through the internal USB header
2 x USB 2.0/1.1 ports on the back panel
Chipset+2 USB 2.0 Hubs:
8 x USB 2.0/1.1 ports (4 ports on the back panel, 4 ports available through the internal USB headers)

there are many of them which are the same, so I had to use USB Device Tree Viewer to figure out, Now I want to know which CPU controller is which, the top one or the lower one, I assume its the top cuz it has the prioritized PCI Busing which is Number 1, while the other one is PCI Bus 7

also, I used Device Cleanup tool to clean all unused device entries.


Also, Higher DPI causes less input lag than 400 DPI for example? or is it the same since the mouse displacements are the same?
for example

Which is better?
400 DPI and 0.8 ingame OR 1600 DPI and 0.2 ingame?


EDIT: Came across a Topic that says I should use triple digits+ in ingame sensitivity, for example @400dpi - 0.800 or 0.801 if it doesn't work, did I understand this correctly? or I misunderstood something!

<3
Attachments
gafasdaf~2.png
gafasdaf~2.png (41.02 KiB) Viewed 8423 times

Post Reply