Explore BrainMass

Explore BrainMass

    Network Adapter

    Not what you're looking for? Search our solutions OR ask your own Custom question.

    This content was COPIED from BrainMass.com - View the original, and get the already-completed solution here!

    Many of the functions of an adapter(data link or network) related functionality built into network adapter can be performed in software that runs on the node's CPU. What are advantages and disadvantages of moving this functionality from the adapter to the node?

    © BrainMass Inc. brainmass.com March 4, 2021, 7:48 pm ad1c9bdddf

    Solution Preview

    The key bottleneck for high-speed network is network interface due to the low bandwidth of I/O bus and lack of cache of device registers on the network interface card. Also the current network interface card is designed with an interface similar to a disk's interface. Most current network interfaces require application to use operating system calls. The CPU can only access the on-board registers in an in-order and non-speculative way. All these limitations make network interface card a bottleneck for the high performance network.

    An adapter serves as an interface between the host (CPU) and the network link, so it is on the lowest level of network subsystem in the operating system. The adapter can be divided into two parts:
    - the host half and
    - the link half.

    The host half "talks" to the host via I/O bus, while the link half "talks" to the network link media using particular physical layer ...

    Solution Summary

    Advantages and disadvantages of moving data link or network functionality from the network adapter to the CPU are explored.