Does the USB port matter for Keyboard/Mouse?

[TheLostSwede]

I would say can matter, at least if you're unfortunate to have a Genesys Logic USB hub on your board, as they're terrible and sometimes disconnect USB devices for no apparent reason, especially if more than one device is connected to the hub.

 

[LumiLinkSeeker]

Hello

Got an interesting, perhaps stupid question for yall....

Does it matter which USB port you use for a keyboard or a mouse?

I got an X570 mobo with 4 USB2.0 ports, a few 3.0 (5Gbps ports) and 3.1 Gen 1 Service Pack 3 Professional lite edition whateveritscallednow (10Gbps ports)

The connectors on every any mouse and keyboard I have in the house has 4 pins for USB2.0, cause they don't need more bandwidth but whatever.

So: does it actually matter if I use 480Mbps ports, 5Gbps ports or 10Gbps ports for keyboard and mouse? I don't have a lot of USB devices so I got free ports to go around.

Maybe some are CPU attached vs PCH/Chipset attached? Block diagram isn't available, but does it really make a difference?

Microseconds, few milliseconds?

Thanks

I also recommend using USB 2.0 for mice and keyboards, because they don't require high speeds. USB 3.0 is best reserved for devices that require high transmission speeds, such as hard drives and mobile phones.

 

[192kbps]

I can clear this up, paging @Onasi as well.

The important distinction in this regard is USB full-speed and USB high-speed. The former is restricted to a frame size of 1 ms, thus restricting the polling interval to 1 ms (1000 Hz). The latter has a frame size of 125 us, thus allowing up to 8 microframes per frame (8000 Hz). For a frame size of 1 ms, the lowest possible (ideal) click latency is 0.5 ms on average. Hence, all mice achieving sub-0.5 ms latency numbers at 1000 Hz necessarily are USB high-speed mice (such as the OP1 8K, for instance), and for those, the lowest possible click latency is 62.5 us on average. If set to 1000 Hz, the OP1 8K will send sensor motion data every 1 ms, but if there are any click events, they'll be sent at the next available microframe, regardless of whether motion data would be sent at some other point in time. This way, click latency will be the same regardless of the set polling rate, and as low as possible on a USB high-speed mouse.

Keep in mind that this is part of firmware design, and other USB high-speed mice handle things differently—Razer USB high-speed mice, for instance, do not decouple motion and click data, and the same goes for many USB full-speed mice as well, such as the Logitech G203, which has higher click latency the lower the polling rate. On the other hand, we have the CHERRY XTRFY M64/M68 Pro, which have the same click latency regardless of polling rate, even when wireless. It really is a matter of firmware design.

The lowest delay is 1us, the lowest average delay is 62.5us?

 

[pzogel]

The lowest delay is 1us, the lowest average delay is 62.5us?

Lowest would be 0 us, but that is purely theoretical. Hence, if we were to assume a range of 1-125, the average would be 63 us. Ultimately, none of these values could be achieved in reality. The lowest I've seen so far were 90 us on the Viper 8K/DeathAdder V3.

 

[192kbps]

Lowest would be 0 us, but that is purely theoretical. Hence, if we were to assume a range of 1-125, the average would be 63 us. Ultimately, none of these values could be achieved in reality. The lowest I've seen so far were 90 us on the Viper 8K/DeathAdder V3.

Don’t you agree with the OP1 8K test results on XLAT? Is this a problem with XLAT or XLAT WIKI?

 

[pzogel]

Don’t you agree with the OP1 8K test results on XLAT? Is this a problem with XLAT or XLAT WIKI?

OP1 8K is at 77 us w/ XLAT, which is still theoretically possible at least. In general, I consider XLAT inferior to LDAT, for reasons explained here: https://www.techpowerup.com/review/finalmouse-ultralightx/8.html

 

[GerKNG]

the only real limitation is the power specification of the USB Port. I used a lot of different USB peripherals with polling rates up to 8KHz and they always functioned identically on all motherboards i've ever used them with, which includes almost the whole 8KHz Logitech Lineup, wireless devices, razer mice and keyboards on motherboards from the sandybridge era up to AM5/Raptorlake Refresh.

i've never seen a device working properly at 8KHz polling rates on any USB Port no matter if 20Gb/s USB C ports or all the way down to 2.0 type A... (mostly due to game engine problems and not specifically windows or the Hardware itself) but 4KHz Mice, Keyboards and wireless devices work exactly the same on USB 2.0 as on they would on USB 4 Ports.
I currently use a Deathadder V3 and the optical switch Huntsman V2 at 4KHz on two USB 2.0 Ports without any problems.