CONNECT WITH US

iCatch Launches PAISB Sensor Bridge Platform for NVIDIA Holoscan Ecosystem

News highlights

iCatch Technology Introduces PAISB Sensor Bridge Platform for NVIDIA Holoscan Sensor Bridge Ecosystem. Credit: iCatch

iCatch Technology, a leading provider of AI imaging SoCs and vision processing solutions, today introduced PAISB (Physical AI Sensor Bridge), an HSB-aligned ASIC sensor bridge platform designed to enable scalable multi-sensor data ingest for NVIDIA Holoscan-based Physical AI systems.

The first product in the platform family, PAISB-E1, will be demonstrated at GTC 2026 running on NVIDIA Thor at ApproPho Booth #3420.

As Physical AI systems increasingly rely on multiple high-bandwidth sensors - including cameras, stereo vision modules, and other perception devices - efficient and deterministic data pipelines from sensors to GPU compute platforms have become critical infrastructure. PAISB addresses this challenge by providing an ASIC-based bridge optimized for time-aligned, low-latency sensor streaming into GPU computing environments.

Aligned with the NVIDIA Holoscan Sensor Bridge (HSB) architecture, PAISB enables developers to ingest synchronized sensor streams directly into GPU memory, allowing real-time perception workloads to scale across a wide range of Physical AI applications.

At GTC 2026, iCatch will demonstrate PAISB-E1 using FoundationStereo Depth as an application example. The system streams synchronized multi-sensor data to NVIDIA Thor, enabling real-time perception pipelines for robotics and intelligent machines.

"Physical AI systems demand scalable and deterministic sensor-to-GPU pipelines to process massive volumes of multimodal data in real time," said Weber Hsu, President of iCatch Technology. "With PAISB, we are introducing an ASIC-based sensor bridge platform aligned with the NVIDIA Holoscan ecosystem, enabling developers to build high-performance, time-synchronized sensor architectures for Physical AI platforms."

The PAISB platform is designed to support a broad range of Physical AI applications, including robotics, industrial automation, medical imaging, broadcast infrastructure, and automotive perception systems. By enabling high-bandwidth, synchronized sensor ingestion into GPU-based computing environments, PAISB accelerates the deployment of AI systems requiring precise perception and real-time decision-making.

To learn more about PAISB and explore collaboration opportunities, visit us at ApproPho Booth #3420 during GTC 2026, or schedule a meeting in advance.