Netcope Session Filter offloads predefined types of traffic in hardware and thus it unlocks the possibility to build a wire-speed solution for 100G traffic processing. The deluge of incoming data will no longer be processed in CPU alone.
Abstract: NPUs equipped with a variety of functions and high aggregate throughput serve as an ideal tool for traffic processing on high-speed links. CPUs can perform virtually any network function, since any network function can theoretically be described in software. However, neither CPUs nor NPUs do not offer sufficient performance to guarantee wire-speed traffic processing on 100G links when it comes to more complex tasks such as pattern matching, encryption or longest prefix match. This whitepaper discusses why FPGA’s are ideal for acceleration of NPUs and CPUs running VNFs, because they leverage programmable data plane with wire-speed throughput and deterministic processing. Several specific use cases of Netcope Session Filter, specialized firmware designed for traffic offloading in FPGA PCI Express network card, are described in detail in the latter part of this whitepaper.
Download the whitepaper here.
If you wish to learn more about Netcope Session Filter, visit our Product Page or download a Product Brief.