Holoscan Framework/Basics and Foundations/Ecosystem

Holoscan Framework
Basics and Foundations
Low Latency Capture Holoscan Ecosystem
Holoscan Sensor Bridge
Architecture Hardware Connection Bring Up Process Running Demo Performance Applications
DPDK Framework
Introduction Installation and Testing
Contact Us

Holoscan Framework

NVIDIA's Holoscan is a comprehensive platform designed to accelerate the development and deployment of AI and high-performance computing (HPC) applications for real-time insights. It's particularly well-suited for industries like healthcare, automotive, and robotics, where real-time processing of vast amounts of data is crucial.

Key Components of Holoscan:

• Hardware: Holoscan leverages NVIDIA's powerful GPUs and other specialized hardware, like FPGAs, to handle intensive computational tasks.
• Software: The Holoscan SDK provides a robust framework for developing and deploying AI applications. It includes tools for data ingestion, preprocessing, model inference, and visualization. It also supports Python, C++ and Graph Composer.

Holoscan Sensor Bridge

The Holoscan Sensor Bridge extends the capabilities of the Holoscan Framework by serving as a versatile interface for integrating diverse sensors into real-time AI applications. Sensors such as cameras, LiDAR, and medical imaging devices produce massive amounts of data that must be ingested, synchronized, and processed with minimal delay. The Sensor Bridge simplifies this process, acting as the critical link between hardware sensors and AI-driven insights. Capabilities of the Holoscan Sensor Bridge:

• Sensor Integration: Supports a wide variety of sensors, including high-definition cameras, ultrasound probes, and industrial LiDAR devices. Provides out-of-the-box support for popular data transfer protocols like GMSL2, GMSL3, and Ethernet-based sensors. At the moment, it is experimental and support cameras only.

• Low-Latency Data Ingestion: Ingests high-bandwidth data streams in real-time, preserving the temporal accuracy needed for low-latency applications such as AR overlays or surgical guidance. It uses DPDK-compatible ethernet for better latency.

• Data Synchronization and Processing: Synchronizes multi-sensor inputs to ensure coherent data streams, even in complex setups with multiple devices. Allows for pre-processing operations, such as filtering, normalization, or compression, before feeding data into AI inference pipelines.

• Edge Scalability: Optimized to run efficiently on NVIDIA edge platforms like the Jetson Orin and IGX Orin. It can scale from single-sensor setups to large, distributed systems with multiple inputs.

Hardware Composition of the Holoscan Sensor Bridge

RidgeRun Services

RidgeRun has expertise in offloading processing algorithms using FPGAs, from Image Signal Processing to AI offloading. Our services include:

• Algorithm Acceleration using FPGAs.
• Image Signal Processing IP Cores.
• Linux Device Drivers.
• Low Power AI Acceleration using FPGAs.
• Accelerated C++ Applications.

And it includes much more. Contact us at https://www.ridgerun.com/contact.