Netronome Heterogeneous Reference Architecture
This architecture couples the Netronome Network Flow Processor (NFP-32xx), optimized for L2-L7 processing, with one or more general-purpose multicore x86 CPUs. Designs based on this architecture are powered by the NFP’s 40 multi-threaded programmable networking cores, running at 1.4GHz to deliver over 56 billion instructions per second with 320 hardware threads to optimize memory utilization -- allowing for 1800 instructions per packet at 30 million packets per second. This enables 20Gbps of L2-L7 deep packet processing with line-rate security and I/O virtualization for millions of simultaneous flows. A virtualized PCIe Gen2 interface on the NFP provides a high-throughput datapath between the NFP and multicore x86 CPUs to allow equipment providers to deliver high-performance, flexible and field-programmable systems for applications that meet the stringent requirements that today’s network services require.
