Interrupt conflict between IDE driver and Network driver identified as problem

When we did our reviews of the Gigabyte K8NXP-SLI board and DFI's LanParty UT nF4 Ultra-D boards we noticed unusually high processor usage when testing the integrated network card on the nForce 4 chipset.  Perplexed we contacted NVIDIA about this issue and we have spent a couple weeks on trying to identify and resolve this issue with them and we finally have an answer now. It turns out that the IDE driver and the network driver are sharing the same interrupt and this was causing some issues with Active Armor - we have a couple work arounds from NVIDIA. Here's a copy/paste of the correspondence:

1- Remove the NVIDIA IDE driver and use the Microsoft IDE driver. Once that is done, ActiveArmor works again as intended. CPU utilization drops from about 40% to 20%.

2- Unplug the SATA drive and try different ports on the motherboard, keep enabling and disabling the IDE driver till you come to a state where the IDE driver and the network driver don't share the same interrupt... Once that is done, ActiveArmor works as intended.

NVIDIA is looking for a better solution right now and have both their BIOS and IDE teams working on it to find a permanent fix.  This may be vendor specific BIOS problem as it does not show up on the NVIDIA reference solution but since we have a limited number of consumer boards on hand at the moment we cannot verify this.  Big thanks to the guys over at NVIDIA as they were very persistent and have spent countless hours on the phone with us in trying to resolve this!