Gidel at ChipEx 2025

ChipEx 2025: Gidel Demonstrated Real-Time FPGA Imaging and Edge AI

At the ChipEx 2025 Exhibition, Gidel demonstrated FPGA-based imaging technologies for high-performance vision, embedded AI, and real-time image-processing applications.

The live demonstration featured FantoVision20-GigE, combining NVIDIA Jetson processing with FPGA-based 10 GigE Vision acquisition and real-time image processing in a compact Edge AI platform.

FantoVision20-GigE: Real-Time 10 GigE Vision Demo

The FPGA handled deterministic image acquisition, HDR, image enhancement, and Quality+ Compression before transferring the processed image data to the Jetson for AI and application processing.

By moving these pixel-intensive operations to the FPGA, the system reduced acquisition and image-processing overhead on the Jetson, leaving more CPU/GPU resources available for AI inference and application processing.

Key capabilities demonstrated at ChipEx 2025 included:

• Real-time 10 GigE Vision acquisition for high-bandwidth imaging
• Single-frame HDR processing to improve image detail across challenging lighting conditions
• Quality+ Compression for substantial data reduction while maintaining high image quality
• Deterministic FPGA processing for predictable real-time operation
• Sub-frame processing latency through FPGA-based acquisition and inline image processing

Engineers evaluating GigE Vision camera selection can also review Gidel’s guide to GigE Vision cameras to compare bandwidth, connectivity, cabling, synchronization, and acquisition requirements.

FPGA Processing for High-Performance Vision Systems

The ChipEx 2025 demonstration showed how FPGA processing can be integrated directly into the image-acquisition pipeline.

Instead of transferring raw camera data to the NVIDIA Jetson and relying on the CPU/GPU for every processing stage, the FPGA can handle acquisition, protocol processing, HDR, image enhancement, compression, and data preparation before the image reaches the Jetson.

This architecture helps maintain predictable low-latency operation while preserving more Jetson computing resources for AI inference and application processing.

For semiconductor and industrial vision systems, embedded AI, and other high-performance imaging applications, this approach can be particularly valuable when high-resolution sensors, high data rates, deterministic acquisition, and real-time processing must operate together.

High-Bandwidth FPGA Imaging Interfaces

Gidel’s FPGA imaging platforms support multiple camera interfaces for high-performance acquisition and processing:

• GigE Vision frame grabbers for flexible high-bandwidth Ethernet-based acquisition
• CoaXPress frame grabbers for very high-bandwidth acquisition, including CXP-12
• Camera Link frame grabber for deterministic acquisition with established industrial camera interfaces

These FPGA-based platforms can combine camera acquisition with real-time processing, image enhancement, compression, and data preparation within the same processing pipeline.

Explore Gidel FPGA Imaging Solutions

• FantoVision Edge AI Systems
• Real-Time HDR
• Quality+ Compression
• Gidel PCIe Frame Grabbers

ChipEx 2025 Event Details

📆 Tuesday, May 13, 2025
📌 EXPO Tel Aviv

ChipEx 2025 brought together semiconductor, electronics, and technology professionals to explore developments across the chip industry and related high-performance computing applications.

Gidel participated with a live demonstration of FPGA-based image acquisition, real-time processing, HDR, compression, and embedded Edge AI.

Official event website: ChipEX