RidgeRun Software Developer Blog

Real-time AI at the edge is reshaping industries from medical devices to broadcast media. NVIDIA’s ecosystem—GStreamer, DeepStream, Holoscan SDK, and Holoscan for Media—offers multiple approaches to building streaming AI pipelines, each with distinct trade-offs in latency, scalability, and architecture. This deep dive clarifies how these frameworks differ and when to use each.

Media over QUIC (MoQ) enables low-latency, multi-track streaming for robotics, edge AI, and embedded systems. Built on QUIC, it supports secure, multiplexed data delivery with no head-of-line blocking. RidgeRun’s Gst-MoQ solution demonstrates scalable publisher–relay–subscriber architecture, dynamic track selection, synchronized telemetry, and real-time multi-camera streaming.

Enable Data Plane Development Kit DPDK on NVIDIA Jetson

Measuring the glass to glass latency of the Holoscan Sensor Bridge and the NVIDIA Jetson AGX Orin interconnected over ethernet

RidgeRun Video Stabilization Libary is the novel IMU-aid video stabilization for embedded systems on drones, robotics and surveillance

Reduce sensor capture latency with the Low Latency Holoscan Sensor Bridge

GStreamer video overlay with RidgeRun QtOverlay, HTML Overlay, and Fast Embedded Overlay with hardware acceleration

HTML Overlay is the latest solution for streaming content overlaying, using HTML webpages in combination with OpenGL for embedded systems.

This blog post showcases the new feature of data channel support in the WebRTCWrapper plugin

CUDA Image Signal Processing (ISP) library: a CUDA-accelerated solution for image preprocessing that offloads to processing to NVIDIA GPUs.

Due to the increment of edge and IoT devices in recent years, opportunities for developers to create new technologies or improve existing...

The number of devices and applications capable to connecting to the internet is growing day by day. Video calls, messages, photos, and...