MegaMatcher Accelerator Extreme
MegaMatcher Accelerator 13.0 Extreme is a family of biometric solutions for fast fingerprint, iris and face matching on the server-side of an AFIS or multi-biometric system. The solutions are intended for national-scale biometric identification projects with hundreds of millions of people enrolled in the database.
These biometric matching engines can be used with MegaMatcher Accelerator 13.0 Extreme:
- Fast fingerprint, iris and face matching engines that can be used separately or together. See technical specifications for engine comparison and licensing model for engine availability.
- Fingerprint, face, iris and voiceprint matching engines that can be used separately or together to validate matching results produced by the fast fingerprint, face and/or iris engines. See MegaMatcher SDK reliability tests for more information.
- MegaMatcher Accelerator software is distributed as Docker containers for using on Linux OS.
Server hardware is optionally available.
In this case the MegaMatcher Accelerator 13.0 Extreme software will be pre-installed by Neurotechnology on each unit, and the customers will receive ready-to use hardware/software solution.
HPE ProLiant DL360 Gen10 server units are offered with these specifications:
- 2 x Intel Xeon Gold 6126 processor (12 cores, 19.25M cache, 2.6 GHz);
RAM, depending on the number of biometric engines:
- 512 GB RAM – for single biometric engine or two biometric engines;
- 1024 GB RAM – for three biometric engines;
- additional 512 GB RAM needed when large fingerprint templates are used.
- 2x HPE 400GB SATA 6G Write Intensive (2.5") SC SSD or HDD with similar capacity;
- HPE iLO Advanced;
- HPE Smart Array P408i-a SR Gen10 (8 Internal Lanes_2GB Cache) 12G.
- Multiple MegaMatcher Accelerator 13.0 Extreme units can be combined using the included cluster software to reach a higher level of performance.
A client communication module is included with MegaMatcher 13.0 Extended SDK. The module allows the sending of a task to MegaMatcher Accelerator, querying the status of the task, retrieving the results and then removing the task. A high-level API is provided for the developer, all low-level communications are hidden.