Sensor processing platform enables fast technology adoption

1 min read

Mercury Systems has unveiled a new Model 8258 development platform based on a dual AMD-Xilinx Versal AI Core device.

Incorporating Mercury’s SCFE6931 signal processing module, this ready-to-run platform has been developed to enable engineers to start application development right “out of the box,” saving them time and reducing project risk.

The Model 8258 platform allows customers to deploy Versal AI Core technology, giving users the ability to react quickly and make real-time tactical decisions that are essential for survivability and mission success.

“There is a growing need for artificial intelligence capabilities in autonomous unmanned vehicles/aircraft and munitions,” said Ken Hermanny, senior director and general manager, Mercury Mixed Signal. “In a world where incremental change is no longer enough, Mercury is helping the A&D industry bend the curve by developing more powerful and secure processing systems, of which our new Model 8258 platform is a prime example.”

“The heterogeneous mix of computing resources inherent to the Versal AI Core series gives designers the freedom to assign computing tasks to the most suitable processing engine, as well as being able to adapt and reassign resources dynamically, as needed.” said Manuel Uhm, Director of Silicon Marketing at AMD. “Mercury’s Model 8258 is a nice addition to the Versal ecosystem by harnessing the resources of two powerful AMD-Xilinx VC1902 adaptive SoCs and simplifying development in a preconfigured platform.”

The 8258 development platform is intended to make it much easier for engineers to unlock the performance needed for advanced defence system solutions found in the SCFE6931 processing module.

The 8258 supports application development on the actual SCFE6931 hardware, versus a lab-only evaluation board, so the IP, software and hardware can transition directly to deployed signals intelligence and electronics intelligence systems, where data can be more quickly analysed and identified.

According to Mercury the platform is suitable for:

  • Developing artificial intelligence applications such as target classification, signal decoding/decryption, image recognition and pattern recognition
  • Navigator FPGA Development Kit (FDK) allows users to develop custom FPGA IP to add application-specific features
  • Navigator Board Support Package (BSP) Includes software functions and libraries to program the board and control the Navigator IP
  • Navigator includes ready to run examples to get started immediately
  • Optional support for four optical 100 GigE interfaces for fast, parallel processing and data throughput
  • Two Versal VC1902 processors on the 6U board doubles the heterogeneous processing resources available