AIR-T Application Development for FPGA
Deepwave’s BitStream product is a development kit for fine-tuning an AIR-T for ultra-low latency RF use cases – where every millisecond counts.
BitStream provides the tools for developers to boost the AIR-T’s analysis capabilities. It allows engineering teams to take advantage of the AIR-T’s unique architecture, and write custom programs to preprocess data on the FPGA chip before data arrives on the GPU for AI-driven analysis.
Consider, for example, airspace safety monitoring for a logistics site that uses drones for package movement. In this use case, it might make sense for the logistics site managers to tune the AIR-T to screen out drones from their fleet so the AIR-T can focus on detecting unknown drones to enhance safety for the operating drone fleet. BitStream would be a perfect tool for this job.
AIR-T optimization efforts, using BitStream, include FPGA-focused:
- Custom signal processing blocks
- Communication physical layers
- Pre-processing for deep learning applications
- Feature extraction
Firmware Development Accelerator for the AIR-T
BitStream speeds AIR-T FGPA firmware development in two ways:
- An intuitive, developer-friendly FPGA interface. The Deepwave team abstracted away all of the complications of writing custom firmware to help engineers interact with the FPGA in a simplified, easy-to-use interface.
- Reusable system capabilities. Board-specific system management – power, clocking, timing, and bus architecture – are handled within the BitStream software which allows developers to utilize existing FPGA computational resources without the long initial development times typical of bringing up an entire system
AIR-T’s Unique Memory Architecture
The figure to the right shows the unique shared memory architecture of the AIR-T product line. With the AIR-T, the GPU and CPU share the same memory which allows signal data to be equally accessible to both compute resources. The signal data is transferred to/from this shared memory using the DMA engine included in BitStream.
Traditional SDR Memory Architecture
This is in contrast to a traditional software defined radio shown to the left. In typical SDR systems, the GPU and CPU are only connected by a PCI express bus requiring an extra memory copy of the signal data to access the GPU compute resource. This extra copy adds significant latency to the application.
The presence of the GPU / CPU shared memory on the AIR-T allows applications to operate with reduced latency while still leveraging the compute power of the GPU. By combining the AIR-T hardware, AirStack Core, and BitStream, Deepwave offers the only SDR with combined FPGA, GPU, and CPU in a fully reconfigurable infrastructure.