HomeElectronics NewsSingle-Board Platform Combines FPGA and AI

Single-Board Platform Combines FPGA and AI

A new single-board platform integrates FPGA-based RF processing with AI acceleration for wireless applications. 

The AntSDR T510 AI

MicroPhase Technology has introduced AntSDR T510 AI, which combines an AMD Zynq UltraScale+ RFSoC ZU47DR processor and an NVIDIA Jetson on a single board for supporting real-time RF processing and artificial intelligence. Designed for next-generation wireless communication systems, the AntSDR platform offers capabilities including signal acquisition, processing, analysis, and artificial intelligence on the same board, reducing the latency associated with multi-board configurations. It provides support for eight synchronized transmission and reception channels, RF coverage from 1 MHz to 6 GHz and AI acceleration capabilities for applications like phased array radar, massive MIMO, spectrum monitoring and intelligent wireless sensing.

The RFSoC is made up of programmable logic, embedded processing units, and high-speed RF data converters on a single piece of silicon. The platform includes eight 14-bit ADC channels operating at up to 5 GSPS, eight 14-bit DAC channels running at up to 9.85 GSPS, and an 8T8R configuration with synchronous ADC/DAC capability. There are eight channels that can handle 2 GHz baseband, whereas the onboard Jetson board is used for AI inference, signal classification, spectral analysis, feature extraction and wireless environment perception.

For larger installations, multiple boards can be synchronised to provide 16, 32, or more RF channels. Timing modes can be provided via external clocks, OCXO/TCXO on the board, GPS time, and PPS trigger capability. This board comes ready to go with Ubuntu 22.04 and CUDA pre-installed and is compatible with GNU Radio, SoapySDR, and IQTAXI driver architecture. The applications can be written using Python or C++ via high-level APIs, making FPGA development skills less necessary.

MicroPhase intended to launch hardware reference designs, firmware source code, host-side streaming examples, and jetson AI workflow examples through its public GitHub repository.

Click here for the original announcement

SHARE YOUR THOUGHTS & COMMENTS

EFY Prime

Unique DIY Projects

Electronics News

Truly Innovative Electronics

Latest DIY Videos

Electronics Components

Electronics Jobs

Calculators For Electronics