Network feature extraction involves the storage and classification of network packet activity. Although primarily employed in network intrusion detection systems, feature extraction is also used to determine various other aspects of a network's behavior such as total trafc and average connection size. Current software methods usedfor extraction of network features fail to meet the performance requirements ofnext-generation high-speed networks. In this paper, we propose an FPGA-based reconfigurable architecture for feature extraction oflarge high-speed networks. Our design makes use ofparallel rows of hash functions and sketch tables in order to process network packets at a very high throughput. We present a detailed description of our architecture and its implementation on aXilinx Virtex-II Pro FPGA board, andprovide cycle-accurate timing results forfeature extraction of input networking benchmark data. Our results demonstrate real-world throughputs of as high as 3.32 Gbps, with speedups reaching 18 x when compared to an equivalent software implementation.