NVIDIA® Jetson™
RidgeRun is proud to support the NVIDIA Jetson family of modules, which enable high compute capabilities, thanks to their powerful GPUs and hardware acceleration units, on a small device. The NVIDIA Jetson platform is ideal for autonomous machines, smart devices, video-based artificial intelligence, autonomous mobile robots, drones,
multi-camera systems and much more.
NVIDIA just released the most powerful SOMs so far!
Make sure you check the newest Jetson AGX Orin™ and the Jetson Orin NX™ modules in their different configurations and bring server-class AI performance to you products.
NVIDIA is a worldwide leading company in terms of technological development in various areas that include:
NVIDIA is well-known for providing powerful hardware modules and diverse software tools for each of the areas where it is involved. Their team provides excellent customer support service in regards to their hardware and software products, which is enhanced with the cooperation of a highly active developers' community.
For the embedded systems area, NVIDIA offers the Jetson family of modules with GPU-accelerated parallel processing, the world’s leading embedded AI computing platform. The Jetson portfolio of devices delivers more performance and features for autonomous machines and other edge AI devices than any other platform available in the market.
About NVIDIA
Supported NVIDIA Platform: Jetson
The NVIDIA Jetson modules contain a set of features that make them suitable for applications like inference on the edge for deep learning and artificial intelligence (AI) applications. The NVIDIA Jetson platform is ideal for applications where the analysis and decision making needs to be done on-site, as well as multimedia applications based in frameworks like GStreamer where capturing and encoding several camera streams is a must.
The NVIDIA Jetson is characterized by high performance, low power, small size and powerful computational capabilities devices for embedded systems. Their powerful GPU, CPU, video convert/encode/decode hardware accelerated units (capable of handling 4K resolution video), multi-camera interface, Deep-Learning hardware accelerated units, and more make this device the best system for embedded learning, computer vision, graphics, and GPU computing.
NVIDIA Jetson Ideal Applications
The Jetson modules enable the design and development of (but not limited to) the following applications:
• Autonomous Machines
• Computer vision
• Artificial Intelligence
• Autonomous Mobile Robots
• Medical Imaging
• Robotics
• Augmented Reality (AR)
• Industrial IoT
• Smart Spaces
• Smart devices
NVIDIA Jetson Software Tools
NVIDIA offers the following software tools for its Jetson platform modules:
NVIDIA JetPack™ SDK: Bundles Jetson platform software including TensorRT, cuDNN, CUDA Toolkit, VisionWorks, GStreamer, and OpenCV, all built on top of L4T with LTS Linux kernel.
NVIDIA® Jetson™ Linux Driver Package (L4T): Provides the Linux kernel, bootloader, NVIDIA drivers, flashing utilities, sample filesystem, and more for the Jetson platform.
NVIDIA DeepStream SDK: A collection of APIs and libraries to help developers build and deploy robust intelligent video analytics (IVA) solutions on the Jetson
platform. it enables building complex applications including image capture, encode, decode, inference using TensorRT, and much more.
NVIDIA Isaac SDK: A collection of libraries, drivers, APIs, and other tools that makes easy to add AI into next-generation robots for perception, navigation,
and manipulation. It facilitates to create and deploy AI-powered robotics.
Products using the NVIDIA Jetson modules - with their 4K60 high-resolution video transform/encode/decode HW accelerated capabilities, their Artificial Intelligence on the edge features and their powerful GPUs - require a highly tuned software stack for maximum performance. Profile your video throughput now.
NVIDIA Jetson Applications
Development Tool: GstDaemon
GStreamer Daemon is gst-launch on steroids created by RidgeRun where you can create a GStreamer pipeline, play, pause, change speed, skip around, and even change element parameter settings all while the pipeline is active. Also called gstd, GStreamer Daemon allows Jetson developers controlling audio and video pipelines using TCP connection messages through the library provided or the example client application.
Development Tool: GstShark
GstShark is an open-source project from Ridgerun that provides benchmarks and profiling tools for GStreamer. It includes tracers for generating debug information plus some tools to analyze it. GstShark provides easy to use and useful tracers to Jetson developers, paired with analysis tools to enable straightforward optimizations. The project leverages GStreamer's tracing hooks and standard tracing and plotting tools to simplify the process of understanding the bottlenecks in your pipeline.
NVIDIA Computer Vision
The NVIDIA Jetson family of modules are world-class for computer vision and embedded learning due to their powerful computational capabilities for embedded systems, deep-learning hardware accelerated units and GPU-accelerated parallel processing. Using NVIDIA Jetson is ideal for companies and development teams working towards developing their own computer vision, AI and Industrial IoT platforms. RidgeRun can help enable your NVIDIA computer vision application development, contact us to get started today.
Ready to work with RidgeRun? Find out more about our
Engineering Services.
RidgeRun Plugins & Drivers For NVIDIA Jetson
Image Stitching for NVIDIA Jetson Orin™
RidgeRun has created an image stitcher that seamlessly merges images from different cameras into a single, wide angle result. The algorithm is accelerated using CUDA so it can keep real time performance even when using 4K images. The overlap between adjacent images is smoothed using a feathering technique that improves the resulting image quality. Combined with our CUDA Undistort, V4L2 Capture Drivers and GstVPI plug-in, you can create a fully-fledged image stitching solution for your product. Check how our Jetson Orin™ Reference Design makes use of GstStitcher.
Custom Linux Drivers: Jetson VL42 Video Capture Drivers
V4L2 is the official Linux Kernel API to handle capture devices like camera sensors, video decoders or FPGAs feeding video frames to the Jetson family of modules, including Jetson Orin™. RidgeRun has more than 15 years of experience creating custom Linux V4L2 drivers for embedded systems. The customer selects the hardware sensor or chip and RidgeRun creates the V4L2 driver. RidgeRun drivers can be used with V4L2 compliant frameworks like GStreamer.
GStreamer CUDA Plug-in: GstCUDA
GstCUDA is a GStreamer plug-in and framework developed by RidgeRun to enable easy integration of CUDA algorithms into GStreamer pipelines on Jetson modules. GstCUDA allows users to easily develop custom GStreamer elements that execute any CUDA algorithm without concerns of memory copies.
Unlike other solutions, the GstCUDA framework includes the source code and it abstracts the complexity of both CUDA and GStreamer so you can focus on your algorithm. Also, it provides out of the box quick prototyping elements, that dynamically loads an external CUDA algorithm library, to speed up the development process and reduce time to market of your products
GStreamer WebRTC Plug-in: GstWebRTC
GstWebRTC is a GStreamer plug-in developed by RidgeRun to convert GStreamer pipelines into WebRTC compliant endpoints. It allows audio and/or video streaming between Jetson modules and PCs using the WebRTC protocol. GstWebRTC supports data, audio and video channels out of the box. Check how our Jetson Orin™ Reference Design makes use of GstWebRTC
GStreamer Pan - Tilt - Zoom Plug-in: GstPTZR
GstPTZR is a GStreamer plug-in that emulates a PTZR (Pan, Tilt, Zoom and Rotate) camera using digital image transformations. They represent physical degrees of freedom in which the camera may rotate in order to capture different parts of the scene. GStreamer GstPTZR uses OpenGL HW accelerated transformations to enable real-time performance on resource-constrained embedded systems.
GStreamer QT Plug-in: GstQtOverlay
GstQtOverlay is a GStreamer plugin that renders QT graphics on top of a video stream. This rendering occurs directly into the GStreamer video frame buffer, rather than in physical graphics memory, with the resultant video having the QT graphics image merged into the video. With the RidgeRun GstQtOverlay element, GUIs will be available not only at the video display but also embedded in file recordings and network streaming, for example.
NVIDIA Jetson Developer Kits and Software Tools
New NVIDIA HW Release! Jetson Orin™
NVIDIA has just released the most powerful family of SOMs yet! The Jetson Orin™ is a set of power-efficient system-on-modules that bring server-class AI performance to the edge. It comes in two versions: AGX Orin™ and Orin NX™ with different memory capabilities each. With up to 275 TOPs the Jetson Orin brings 8X the performance of the Jetson Xavier and the ability to perform multiple simultaneous inferences.
