HPT comes to mind.

I haven't been impressed with the latency using the PLX adapter with Torque.

I'm used to a sample every 200 ms, not every 1000.