top of page

FPGA for DeepLTK

FPGA Acceleration for Deep Neural Networks

Unlock ultra-low latency AI inference on NI FPGA hardware with the DeepLTK FPGA Addon

The DeepLTK FPGA Addon enables seamless deployment of deep learning models to FPGA-based hardware directly from LabVIEW.

 

Built for real-time, high-throughput applications, it allows developers to run neural network inference on NI FPGA targets with up to 100x acceleration compared to CPU-based execution. No HDL coding is required, making it easy to integrate AI into deterministic systems.

Key Features​

Low-Latency Inference

Achieve sub-millisecond inference performance, ideal for real-time and latency-critical applications.

Optimized Resource Utilization

Custom-developed IP cores are designed to minimize logic usage and reduce power consumption on the FPGA.

FPGA Code Generation

Supports all NI FPGA targets with automatic bitfile generation for supported network layers.

Real-Time Ready

Integrates with NI Real-Time OS and hardware (e.g., CompactRIO, sbRIO) for deterministic execution.

Integration & Requirements

The addon integrates seamlessly with DeepLTK and supports deployment to NI FPGA-based platforms such as CompactRIO, sbRIO, and FlexRIO systems.

Installation

Delivered as a VIPM (VI Package Manager) package, including the addon, documentation, and ready-to-run examples.

Requirements

  • DeepLTK v8.0.1 or later

  • LabVIEW 2020 or newer (64-bit recommended)

  • LabVIEW FPGA Module

  • NI-RIO Driver

  • Compatible NI FPGA targets (e.g., cRIO, sbRIO, FlexRIO)

Workflow

1. Design & Train

Create and train neural networks using DeepLTK in LabVIEW.

2. Export

Optimize and export the trained model for FPGA deployment.

3. Compile Bitfile

 Build the bitfile and deploy it to the NI FPGA target.

4. Run

Load the model and perform real-time inference directly on the FPGA.

FPGA Acceleration Toolkit Comes as
Add-On for DeepLTK

DeepLTK

Deep Learning Toolkit for
LabVIEW

A comprehensive, general-purpose toolkit for integrating deep learning into LabVIEW. Fully developed in LabVIEW, it requires no third-party frameworks such as PyTorch or TensorFlow. DeepLTK supports training and deployment of deep neural networks for a variety of tasks, including image classification, object detection, signal pattern recognition, and anomaly detection.​

Download Trial
Learn More

Ready to deploy deep learning on FPGAs?

Run low-latency AI models on CompactRIO, sbRIO, and FlexRIO targets with our FPGA Add-On for DeepLTK.

bottom of page