Hi all, Got something that's been bugging me and thought someone here might know the answer. I recently connected a vhd over iSCSI to my 8.1 workstation running from a 2012 R2 installation that has an Adaptec RAID card in passthrough mode on a dedicated Xeon\ESXi machine. I've got a 2Gb/s LACP team on the workstation and another on the VM. I've been getting transfer speeds however, that I can't really explain; way over 2G wirespeed. In the same respect, the same has been happening on writes from one local drive to another. The local writes can exceed 1GB/s (byte not bits) when it is reading from the Samsung 850 system drive. It can usually maintain this speed on smaller transfers, but when transfers get bigger, say greater than 50-60GB, or I start stringing a lot of operations together, throughput drops by a lot, almost like it is then trying to catch up. I can't quite figure out what is going on. The workstation has a ridiculous amount of memory (64Gb). I figured on the local writes what might be happening is the write request is getting cached to memory and then "promised" a write or something, allowing Windows to show it as a completed operation even if physically it hasn't been written. Just speculation on my part. As for the writes over the network, maybe the iSCSI disk is being treated the same as a local disk? ...Or maybe the iSCSI protocol is able to do something tricky, using all the bandwidth in both directions over the wire. The latter seems unlikely to me. Can anyone explain what is happening here? Normal network read/writes over SMB max out at about 90% of theoretical (~210-220MB/s)... Not really a problem; just had me curious. I don't think it matters, but the network traffic is flowing over a Cisco Catalyst 3750 switch, and the NICs at both ends are Intel. None of the NICs involved support iSCSI offloading or function as converged network adapters. Any insight is appreciated.