Repositories

This document lists the open source repositories containing software known to work well with Deepwave Digital Products, including both third-party software and Deepwave Digital software made available under open source licensing terms. Open source software developed by Deepwave Digital is available via GitHub at https://github.com/deepwavedigital.

GR-Wavelearner

This out of tree (OOT) module for GNU Radio contains code to provide an interface to call NVIDIA's TensorRT deep learning binaries from a GNU Radio flowgraph. TensorRT allows for deep learning networks to be optimized for inference operations on an NVIDIA graphics processing units (GPU).

For an example of how to use GR-Wavelearner, see our presentation here

GitHub Repository URL: https://github.com/deepwavedigital/gr-wavelearner

GR-CUDA

This OOT module contains experimental code on integration of GPU processing into GNU Radio by using the PyCUDA library to run CUDA code from within GNU Radio.

Tutorial: CUDA Blocks with GNU Radio and the AIR-T

GitHub Repository URL: https://github.com/deepwavedigital/gr-cuda

AirStack-Examples

We provide a number of examples on how to use the AIR-T. While these examples are provided with a fresh AirStack installation, they are updated periodically between releases. Additionally, if you upgrade using the .deb files instead of flashing the Jeston image, they are likely not included.

GitHub Repository URL: https://github.com/deepwavedigital/airstack-examples

RAPIDS GitHub

RAPIDS is an NVIDIA open source project to GPU accelearate data science. cuSignal is a GPU accelerated version of scipy.signal.

cuSignal

The RAPIDS cuSignal project leverages CuPy, Numba, and the RAPIDS ecosystem for GPU accelerated signal processing. In some cases, cuSignal is a direct port of Scipy Signal to leverage GPU compute resources via CuPy but also contains Numba CUDA kernels for additional speedups for selected functions. cuSignal achieves its best gains on large signals and compute intensive functions but stresses online processing with zero-copy memory (pinned, mapped) between CPU and GPU.

URL: https://github.com/rapidsai/cusignal

Getting Started Tutorial: cuSignal on the AIR-T


Last update: May 8, 2020