From: "Antti Kurenniemi" 

"Mike N."  wrote in message
news:4apt32h6nthjcon53mgod6vp44974vgdo4{at}4ax.com...
>>Not really. A single thread would still mostly only wait for the disks.
>>Check out the CPU/Memory usage during file copy operations - almost no
>>activity there.
>
>   There is still a benefit to having a write request queued to the target
> drive while having a read request queued to the source drive rather than
> waiting for the data to arrive before passing it on to the target.    Both
> drives would always have an outstanding read or write request rather than
> being idle while the packet is transferred to the other drive.

Yes but that's what the buffer memory in the drive is for, no need to try
to implement a second buffer in the OS level.


>  I have no idea which method XCOPY, etc uses.

Based on the quick test results you posted, I'd say they all use exatly the
same method - just pass the copy request to the OS.


Antti Kurenniemi

--- BBBS/NT v4.01 Flag-5