• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

graphic card & IRQ sharing = performance decrease ??

ShePearl

ShePearl

Limp Gawd
2FA
Joined
Apr 16, 2004
Messages
429
Hi, folks.

Currently,my Radeon 9700 Pro is sharing its IRQ 16 with onboard Intel USB controller - Intel 82801EB USB Host Controller.

I'm using Pentium4C 3.4 GHz (800 MHz FSB)
1 GB Mushkin PC3200 2-2-2 Special (BH-6)
ATi AIW Radeon 9700 Pro (soon to be replaced with NV40 or R420)
Shuttle SB75G2 (i875P Canterwood)
Windows XP Professional Service Pack 1a + latest updates.

Stability is NOT an issue. But I'd like to know if my 9700's (and/or future NV40 or R420) sharing IRQ 16 with Intel 82801EB USB Host Controller decreases any performace ??

I've been out of loops for about 4 years, and back in the old days of Windows 98, I remember sharing IRQ very often leads to stability issues as well as performance decrease.

I'm hearing these days of ACPI motherboard, and hardwares plus Windows XP does proper job of handling IRQ sharings. But what about performance-wise ??

All I'm using in my USB port is an USB mouse, so I can disable USB controllers in BIOS, and use PS/2 mouse instead.

What are your thoughts, folks ?

Sharing IRQ 16 between USB controller and graphic cards = performace decrease ??

Any cooments, opinions, advice would be greatly appreciated.

Cheers.
 
Let me ask you this. Do you have any extra USB ports attached to the system, like in the front of the case or added in where the PCI slot would be?

I had an add in Firewire Card that shared an IRQ slot with my vid card. Since I don't sure Firewire, I disabled it in my Device manager and it got rid of the IRQ sharing. If you do have extra ports attached somewhere, you might have an extra controller showing up in Device manager. Try disabling one in properties and see with controller is running what ports and plug the mouse in the one not shared with the vid card. Also, disable other items you may not use. I got rid of all my shared IRQs this way. I turned off COM ports, LPT ports, Game ports, firewire ports, serial ports and other items I KNOW I don't use.

Unfortunately I don't know too much advanced stuff so I hope someone else will chime in and give you other ideas or a definite answer to your question. I can say that my system felt a little better when I disabled all those worthless ports (for me).
 
...a guess: no performance hit.

It would be a pretty simple test to run; run 3DMark and a couple of other benchmarks with the IRQ sharing enabled. Disable / remove the network card (if it's a card) and run the tests again. Then, you can tell us.....:D

Regards - B.B.S.
 
I can see how this *might* affect performance. I have seen game issues where there were significant performance issues when the video adapter was sharing IRQs with other thingies. However, it's usually limited to bugs in the games. Later patches will resolve these problems, so that makes me thing it's a code issue, not a system performance issue.

Of course the only way to know is to test.
 
Thanks for your inputs. Very much appreciated.

Anyway, I've just tested 3DMark2003 with Intel USB Host controller enabled (IRQ sharing), and disabled (video card gets it own IRQ 16).

The difference in scores was 7 (seven) points. IRQ shared setup was seven points lower than no sharing.

Yes, I believe this 7 points is negligable. I'm not even sure if the score difference was due to IRQ sharing. Also while benchmarking the USB Host Controller is NOT working is it ? Because I leave my mouse alone unlike when I play games - always move mouse so more work is done in USB Host controller in IRQ 16. :confused:

Also I do not believe synthetic benchmarks such as 3DMark, I don't think they are reliable to measure system performance.

Is there any other more reliable way to test this ?

Cheers.
 
Originally posted by ShePearl
Thanks for your inputs. Very much appreciated.

Anyway, I've just tested 3DMark2003 with Intel USB Host controller enabled (IRQ sharing), and disabled (video card gets it own IRQ 16).

The difference in scores was 7 (seven) points. IRQ shared setup was seven points lower than no sharing.

Yes, I believe this 7 points is negligable.
Click to expand...
...you're welcome :cool:

Yes, I also believe that seven points difference with 3DM is almost random, certanly nothing to get worried about.
Originally posted by ShePearl
Also I do not believe synthetic benchmarks such as 3DMark, I don't think they are reliable to measure system performance.

Is there any other more reliable way to test this ?

Cheers.
Click to expand...
...smart, too...:D

You could use any number of the in-game benchmarks available. I'm not up on what's current, but for sure someone on this site is.

Regards - B.B.S.
 
You must log in or register to reply here.
Back
Top