04-28-2006 05:27 AM
04-28-2006 05:12 PM
I will give it a shot since no one replied today. It is generally better to read the IO in a lock instead of one by one. This condition is known as IO bound because the performance is being limited by the IO. Assume you have 16 channels but you only need 6 of them. It would usually be faster to read the 16 channels as a block and index out the data you need than making 6 separate reads. I don't use cFP so speaking in general terms.
Matt