Originally Posted by BlueMonster
You actually gave me six
rep points, so thank you
Overall, I would advise you not to modify these settings away from their defaults because that can sometimes cause weird symptoms which may not be obvious and easy to troubleshoot. Instead, if you think there's a problem with your machine's networking, perhaps you could explain what you're observing and someone may be able to help. Otherwise, if you simply want a speed boost, a better ISP or a higher-performing internet plan is pretty much the only way to fo.
As for the settings themselves...
Interrupt moderation - whenever the NIC hardware notices that it has received an inbound packet, it has to invoke the OS and driver
code for dealing with the situation. The way it does that is by issuing an "interrupt" which literally diverts the system (very briefly) from whatever else it may have been doing and causes the packet reception routines to run. If the machine is receiving hundreds or even thousands of packets per second - which is by no means abnormal - the combined processor hit can be substantial, so "interrupt moderation" causes the NIC to hold its interrupts and batch them together once it has received a certain number of packets, at its own discretion. Disabling moderation causes each incoming packet to immediately trigger an interrupt. That can have a positive effect on the speed of network transfers, but it also means sacrificing processor cycles which would otherwise be spent elsewhere. (To be honest, Realtek is like the McDonalds of NIC vendors, and it's entirely possible that your NIC is not capable of moderation in the first place.)
TCP Checksum Offload - every packet that's leaving your machine has fields for "checksum" values in several of the protocol headers. Think of it as parity information. When the recipient on the other end of the network link receives packets from your machine, the first thing it does is to run the same checksum algorithm over the packet contents, and then compare the result to the info in the packet's "checksum" field. If the two don't match, it concludes that the packet has been corrupted en route, and the packet is unceremoniously dumped. Hence, calculating the correct checksum for each outgoing packet is an important task. It's also a computationally expensive task, in the sense that the OS/processor has to do a fair amount of work to calculate one or more checksums (on different protocol layers) for each outgoing packet. That's where "checksum offload" steps in - it leaves it up to the NIC hardware to calculate outgoing packet checksums, thus freeing up the processor(s) for other tasks.
IP checksum offload - same thing, different protocol layer.
UDP checksum offload - same again.
Note that these settings are in the Device Manager and not in the "networking" control applet. That's because they control low-level NIC behaviour and not its networking stack settins.