Lightning Talk - Deepwave Digital: Artificial Intelligence Radio Transceiver with ...
summary
A compact software‑defined radio with an on‑board NVIDIA Jetson SoM, FPGA and Analog Devices RF chip processes about 125 MS/s per I/Q yielding ~100 MHz usable bandwidth, draws ~15 W nominal (25 W startup), and avoids up to ~50 ms bus‑copy latency by sharing CPU/GPU memory. Demo neural nets classify radar down to ~−10 dB SNR and run natively for deployed TRL7 outdoor CBRS sensors (Keybridge boxes with GPS discipline) to detect Navy radar; also runs GMRS FM demodulation into on‑device speech‑to‑text via CUDA.
AI-written summary · about
related talks
14mThe Synergy of LoRaWAN and Artificial Intelligence
57mPervasive and Sustainable AI with Adaptive Computing
23mDeep learning inference in GNU Radio with ONNX
01h 00mDrones, AI & IOT - What's all the buzz about ?
10mProfiling Whisper AI Model on RISC-V: CPU, GPU, and NPU Performance on the DC-ROMA AI PC
17mYour realtime AI is ngmi
07mKeynote: Modernizing the Development of AI IoT Devices with WebAssembly-based...
34mVoice-Controlled Lights on a Raspberry Pi