Technology1 min read

Enabling Real-Time AI for Scientific Instruments with GPU-Native Data Acquisition

By · Published by Everything Blog

In short

NVIDIA developed DAQIRI, a library integrated into its Holoscan platform, to stream high-bandwidth sensor feeds straight into GPU memory. When tested alongside Lawrence Berkeley National Laboratory's Molecular Foundry, the framework processed over 500 gigabits per second from an advanced electron microscopy camera. The pipeline successfully reconstructed energy-loss spectra and maintained submicron spatial resolution on the fly from 5-micrometer detector pixels. These capabilities enable operators to inspect live data streams and dynamically adapt acquisition regions of interest.

Key points

  • Direct routing of instrument measurements into GPU memory eliminates standard storage bot…: Direct routing of instrument measurements into GPU memory eliminates standard storage bottlenecks during intensive laboratory research.
  • Integrated hardware packet reordering and low-latency network polling ensure steady throu…: Integrated hardware packet reordering and low-latency network polling ensure steady throughput for time-sensitive computing tasks.
  • Real-time spectral processing allows researchers to immediately fine-tune microscope sett…: Real-time spectral processing allows researchers to immediately fine-tune microscope settings and adjust experimental parameters during active experiments.

Sources checked by our analysis pipeline: 1

In science and industry, analyzing sensor measurements in real time can enable AI to adjust processes, guide experiments, and search for better operating conditions. Yet limits on how quickly the measured data can be moved and processed often shape how an instrument is designed, from readout electronics to acquisition timing. What could scientific instrument makers simplify—or make possible—if processing could keep pace with their sensors?

NVIDIA DAQIRI is a software-defined data acquisition library that moves high-bandwidth instrument and sensor data into GPU-ready workflows. It enables developers to process measurements during acquisition without first writing them to disk. An NVIDIA Holoscan platform library, DAQIRI bridges sensors optimized for measurement with programmable GPU platforms for scientific analysis, AI inference, and feedback applications.

For existing Ethernet-capable instruments, this opens opportunities to add processing and feedback capabilities, as described in the previous post, Enable Real-Time AI for High-Speed Data Acquisition with NVIDIA DAQIRI. For new instruments, it allows opportunities for reconsidering acquisition constraints from the start.

This post explores how a new electron microscopy camera and GPU processing enable analysis during acquisition. It then…

Technology