I did a quick capture while the problem was happening on my system. First, I was warned that a lot of events were missed -- I assume that is because the Interrupts/DPCs are at 90% and there is just not enough bandwidth to get everything onto the disk. I know there are things I can do like increase buffers etc that might help, but I have not messed with that yet. What I see is that DPC cpu usage is up and down while the problem is happening and maxes out at about 5%. The interrupt CPU usage is pegged while the problem is happening. The max display is 75% so I assume that CPU usage is higher than that. While this event is happening, there are somewhat higher "hard faults" than at other times, but it maxes out at 9 counts and generally is much lower than that. I didn't look further into the interrupts, but I assume they are USB interrupts since the problem goes away when I disable the USB. Does this sound like a hardware problem? If so, I wounder why if I disable the USB the problem goes away, but it does not come back again when I re-enable the USB. However, when I reboot, the problem is back.