High mouse polling rate = cpu spikes?

[steiner666]

sorry, there aren't any mouse categories to start this thread under, so I'm posting here.

I recently got a new mouse. I had been using a Razer diamondback for about 3 years, and didnt have any problems with it until it started double-clicking on its own and a break in the cord made it unusable. My new mouse is a Logitech G9. I like it, it's pretty smooth.

Anyways, in the software that comes with the mouse you can adjust a lot of different things to fine-tune the mouse. I upped the polling rate to 1000 (1ms) and applied and closed the software and moved the mouse around to test it out, to see if i could tell a difference. It did seem to be a bit smoother, but I noticed that one of my cores would spike to almost full usage even 10-12 seconds when i was continuously moving the mouse around. I stopped moving the mouse and my cpu idled (a steady <1-2% usage), when i started moving it, my cpu usage again started spiking.

I set the polling rate back down to 500 and again moved my mouse around continuously. Again i got cpu spikes, at the same frequency, although this time the usage during the spikes was about halved.

I've read that higher polling rates on mouse causes more cpu usage, but I didn't think it would be such a huge drastic difference, especially now days on a quad core running at nearly 4ghz. I also dont remember seeing any such usage spikes with my old razer, although i dont remember specifically adjusting the polling rate.

The usage is obviously caused by mouse movements, but could there be a problem in the OS or elsewhere in the system that causes these massive spikes? I'd like to have the polling rate on 1000, as i actually do notice a difference. The cpu spikes dont really matter whenever i'm not in a game.... but then again, neither does the polling rate really...

system:
intel c2q q9550 @ 3.8ghz
asus p5e-vm hdmi
gskill 4x2GB ddr2-1000
nvidia gtx260
creative x-fi
2x250gb 7,200rpm WD raid-0
1x1TB WD green
antec neopower 650W
vista ultimate sp2

 

[jsc]

Strange. A 1 ms polling rate is glacially slow compared to CPU core speeds. Are you sure there isn't another factor working?

 

[steiner666]

thats what im thinking. but I havent been able to find anything. it gets worse the higher i set the polling rate, and only happens if i move my mouse around. I thought maybe i was dragging my cursor over a desktop icon or sidebar and maybe the aero visual effects were causing it (not really likely with a beastly GPU), but it happens when i'm just moving my mouse around on blank screen space. Of course i thought it might be from a backgroup app, but i closed everything out, and, like i said, if i just stop moving my mouse the spikes stop also. If i move my mouse around only a little, and slowly, there is either no spike, or a significantly smaller one.

 

[Keyrlis]

I'm having similar issues, but I believe mine is related only to my firefox browser running, as it happens with a PS/2 mouse or my USB trackball.
However, I have found a few sites commenting that the USB port (or controller) may be failing due to a physical problem, even (YIKES!) dust. Have you tried a different port, maybe one connected to a different controller to see if there is any noticeable change?
Also, the software included with Logitech devices is usually a bit bloaty, IMHO, so maybe try it with just the stock windows drivers, then using a program like USB Mouserate switcher to up the polling speed?

 

[HEXiT]

the only time you may have an issue with the mouse causing cpu spikes is if its set with the wrong priority in windows. if its set to high it can interfere with windows you may find that your mouse manufacturer has updated its firmware if its a known problem so it may be worth going to there site and seein if theres an update for either the driver or the mouse itself.