We are excited to announce a transformative evolution for LabVIEW: it is becoming a powerful graph editor. Leveraging ONNX Runtime, LabVIEW now enables engineers to visually orchestrate computational graphs, bringing unprecedented flexibility and efficiency across diverse hardware platforms.
This significant advancement is made possible by our deep integration and active participation in the ONNX Runtime project through Graiphic. Our dedicated teams have mastered this graph-based technology, positioning SOTA to progressively adopt it across all its toolkit offerings.
Initial benchmarks demonstrate compelling results: using our new graph orchestration capabilities, computational functions—such as common computer vision operations—are executed up to 40% faster compared to equivalent OpenCV benchmarks (measured CPU-to-CPU to ensure accurate and fair comparisons). This indicates enormous potential not just in computer vision, but across a wide spectrum of computationally intensive applications.
The LabVIEW Accelerator Toolkit, which we aim to launch next month, marks our first major step in this new direction. It opens the door to real-time signal processing tasks optimized specifically for targeted hardware, leveraging the efficiency and adaptability inherent to graph computing.
Moreover, this graph-based approach provides unique insights into executing code efficiently and optimally across any hardware platform. Going forward, we plan to further enhance ONNX Runtime compatibility with an even broader range of hardware, significantly expanding the possibilities for efficient computation.
Stay tuned for further updates as SOTA continues to integrate this efficient, versatile graph technology across our entire suite of toolkits, pushing the boundaries of performance and capability for engineers and scientists worldwide.
