The Hyperion F1 10G RegEx File Scan instance provides a preloaded IP image of the 10Gbs Regular eXpression Processor (RXP) on the FPGA and a pre-installed SDK allowing the evaluation of Titan IP technology and development of applications to drive the Hyperion F1 10G RegEx File Scan. The SDK is built on top of Amazon Linux and includes the RXP Compiler, an API, a job generator, and C reference applications. The reference applications provided are Hyperion F1 10G RegEx File Scan Reference Application (HRA), HRA lite, and HRA file scan.
The API calls control the start, stop, push data and receive match information from the Hyperion F1 10G RegEx File Scan. The customers application will slice the file into sections, these sections of data, known as jobs, are then sent to the RXP. The solution has been optimized for file scanning with job lengths of no less than 8Kbytes and maximum length of 16Kbytes. The Hyperion F1 10G RegEx File Scan will scan the data against a set of regular expression rules and will report back in an ungreedy mode any matches found. The response comprises of the job ID, the rule ID it matched against, the start pointer and the length of match.
The RXP throughput performance is dictated by the complexity of the rules, number of the rules, number of matches in the data, file access speed and job length. Typical performance for string matching and data analytical RegEx rules is up to 10 Gbps.