Microchip to Boost Edge AI with NVIDIA Holoscan

-

 The NVIDIA Holoscan AI sensor processing platform, SDK and development ecosystem has helped to streamline the design and deployment of AI and high-performance computing (HPC) applications at the edge for real-time insights. Now, FPGAs are unlocking new edge-to-cloud applications for this advanced AI platform while enabling AI/ML inferencing and facilitating the adoption of AI in the medical, industrial and automotive markets.  Microchip's new PolarFire FPGA Ethernet Sensor Bridge is empowering developers to create innovative, real-time solutions with NVIDIA’s edge AI and robotics platforms that will revolutionize sensor interfaces across a wide range of powerful applications.

To enable developers building artificial intelligence (AI)-driven sensor processing systems, Microchip Technology has released its PolarFire FPGA Ethernet Sensor Bridge that works with the NVIDIA Holoscan sensor processing platform. 




Accelerating Real-time Edge AI with NVIDIA Holoscan

The PolarFire FPGA Ethernet Sensor Bridge has joined the NVIDIA Holoscan ecosystem to bridge real-time sensor data to NVIDIA Holoscan and the NVIDIA Jetson platforms for edge AI and robotics. The bridge solution efficiently manages high-bandwidth data from diverse sensors over Ethernet, enabling real-time, high-performance edge AI processing on NVIDIA AI platforms. The power-efficient design is also conducive for small-footprint and energy- or cost-sensitive applications.

PolarFire FPGAs address security concerns in sensor applications by providing embedded security and safety features to help protect against potential cyber threats and provide physical, device, design and data integrity. They are additionally designed with single event upset (SEU) immunity, making them highly reliable in environments subject to radiation, such as space or high-altitude applications and medical environments. The SEU immunity also helps reduce the risk of data corruption and system failures.

PolarFire FPGAs enable multi-protocol support, and this first solution to be released as part of Microchip’s platform is compatible with MIPI CSI-2-based sensors and the MIPI D-PHY physical layer. Future solutions will support a wide range of sensors with different interfaces, including SLVS-EC 2.0, 12G SDI, CoaXPress 2.0 and JESD204B. The platform allows designers to leverage the power of the NVIDIA Holoscan ecosystem while taking advantage of the PolarFire FPGA’s power-efficient technology with low-latency communication and multi-protocol sensor support. 


PolarFire FPGA Ethernet Sensor Bridge

The PolarFire FPGA Ethernet Sensor Bridge accelerates the development of edge AI applications using the NVIDIA Jetson AGX Orin and IGX Orin developer platforms. Based on the power-efficient PolarFire FPGA architecture, this scalable and future-proof platform reduces costs and complexity while addressing the developer’s need for connectivity to a diverse variety of sensors and interfaces. It also supports low-latency interfacing for next-generation edge AI applications in healthcare, robotics and embedded vision, as well as enhanced reliability and embedded security to meet the needs of medical imaging, industrial automation and autonomous systems.




Key features of PolarFire

  • Multi-Protocol Support: The bridge supports MIPI CSI-2 today, with protocols like CoaXPress, SLVS-EC, SDI and JESD204B under development and coming soon.
  • High-Speed Data Transfer: Dual 10G SFP+ Ethernet ports ensure fast data transmission over long distances for AI/ML inferencing on NVIDIA Jetson AGX Orin and IGX Orin platforms.
  • Low Power Consumption: Excellent for power-sensitive applications.
  •  Embedded Security: Hardware encryption, secure boot and more ensure secure, reliable connections between sensors and processors.
  • Seamless NVIDIA Integration: Fully compatible with the NVIDIA Holoscan SDK, enabling streamlined AI deployment using modular components, libraries and pre-built AI models.
  • Data Conversion and Transfer: Sensor data—such as that from an included Arducam MIPI CSI-2 camera—is converted into 10G Ethernet packets compatible with NVIDIA’s Hololink interface, which then transfers the data directly to the Jetson AGX Orin or IGX Orin developer kits for AI processing.
  • Holoscan Integration: You can use the NVIDIA Holoscan SDK to deploy neural networks for AI inferencing and take advantage of prebuilt operators for sensor data processing. The PolarFire FPGA Ethernet Sensor Bridge can send real-time data to Holoscan-enabled devices–crucial for applications like surgical navigation and automated industrial inspections that require instantaneous insights.
  • Plug-and-Play Compatibility: The PolarFire FPGA Ethernet Sensor Bridge’s integration with NVIDIA devices enables a seamless solution for edge-to-cloud applications, allowing for quick sensor connection, configuration and deployment without the need for extensive customization.



Security and Efficiency 

The NVIDIA Holoscan Sensor Bridge handles high-bandwidth data from many sensors via Ethernet very well, thanks to Microchip’s PolarFire FPGA technology’s low power consumption. This makes it possible for NVIDIA AI platforms to do real-time, high-performance edge AI processing. The energy-efficient design is suitable for applications with limited space and those sensitive to energy or cost constraints.

PolarFire FPGAs mitigate security issues in sensor applications by offering integrated security and safety capabilities to safeguard against potential cyber threats and ensure physical, device, design, and data integrity. They are also engineered with single event upset (SEU) protection, rendering them exceptionally reliable in radiation-prone areas, including space, high-altitude applications, and medical settings. The SEU immunity additionally mitigates the danger of data corruption and system malfunctions.


Comments

Post a Comment

For any correction, doubts or suggestion, please let me know

Popular posts from this blog

Registers: The Backbone of Computer Memory

-
Image
 In computer science, a register is an important component of digital devices that stores data and instructions for quick processing. A register is a temporary storage area built into a CPU. Some registers are used internally and cannot be accessed outside the processor, while others are user-accessible.  A register may hold an instruction, a storage address, or any kind of data (such as a bit sequence or individual characters). Some instructions specify registers as part of the instruction. For example, an instruction may specify that the contents of two defined registers be added together and then placed in a specified register. A register must be large enough to hold an instruction - for example, in a 64-bit computer, a register must be 64 bits in length. In some computer designs, there are smaller registers - for example, half-registers - for shorter instructions. Depending on the processor design and language rules, registers may be numbered or have arbitrary names. A pro...
Read more

Preprocessor Directives in C

-
Image
 Preprocessing is an initial phase to process text before compilation. Preprocessor directives are lines of the source file where the first non-whitespace character is #, which distinguishes them from other lines of text. It operates under the control of preprocessor directive which is placed in the source program before the main(). Before the source code is passed through the compiler, it is examined by the preprocessor for any preprocessor directives. The preprocessor directives are always preceded by a hash sign (#). The preprocessor is executed before the actual compilation of program code begins.  Therefore, the preprocessor expands all the directives and take the corresponding actions before any code is generated by the program statements. No semicolon (;) can be placed at the end of a preprocessor directive.  Advantages of preprocessor directives The advantages of using preprocessor directives in a C program include: Program becomes readable and easy to understand...
Read more

Function Generator

-
Image
 In electrical and electronics engineering, a function generator is usually a piece of electronic test equipment or software used to generate different types of electrical waveforms over a wide range of frequencies. Some of the most common waveforms produced by the function generator are the sine wave, square wave, triangular wave and sawtooth shapes. These waveforms can be either repetitive or single-shot (which requires an internal or external trigger source). Another feature included on many function generators is the ability to add a DC offset. Integrated circuits used to generate waveforms may also be described as function generator ICs. In particular it can be made to become a sine wave generator, square wave generator, and triangular wave generator. Also a function generator may be able to vary the characteristics of the waveforms, changing the length of the pulse, i.e. the mark space ratio, or the ramps of the different edges of triangular or sawtooth waveforms. The fu...
Read more