We Are Looking for Beta Testers for SOTA, the Revolution in Deep Learning with LabVIEW!

By Youssef MENJOUR , Graiphic CTO

⏱ 5 minute read · Updated 11:15 AM GMT+0000, Tue January 21, 2025

Graiphic is proud to announce the launch of SOTA in an open beta testing phase. We are actively seeking beta testers to explore this major evolution in the world of deep learning and LabVIEW. This launch represents a complete overhaul of our system, as we rebuilt SOTA from the ground up. The previous generation, known as GIM (Graiphic Installation and Management System), is now replaced by a fully reinvented platform.

Key Innovations in SOTA

New architecture for the LabVIEW Deep Learning Toolkit: We completely rewrote this toolkit, providing greater flexibility and a redesigned architecture to meet modern user needs.
Integration of the universal ONNX format: SOTA introduces a universal file system that simplifies the deployment and compatibility of AI models.
Execution with ONNX Runtime: As an active contributor to the ONNX Runtime project, Graiphic offers optimized integration that enables AI model execution across all hardware platforms, with enhanced performance.

SOTA: A Unified Suite of Tools

SOTA is a unified suite of easy-to-use, interoperable tools—each more innovative than the last—designed for engineers, researchers, and academics. It addresses the fragmentation and complexity of modern AI workflows.

One of SOTA’s major innovations is its adoption of the ONNX Runtime standard, a high-performance and flexible engine for deploying AI models across diverse platforms. Since last month, Graiphic has become the only European company contributing to this open-source, universal major project alongside major American key players like Microsoft, Nvidia, Meta, and Google, among others. This is truly unique.

Another key innovation of SOTA is that its toolkits are built using a graphical programming language based on LabVIEW. LabVIEW is a graphical programming language that enables users to create programs by manipulating visual elements rather than writing text-based code. Unlike conventional flowcharts, which only represent the logical flow of processes without execution capabilities, LabVIEW is specifically designed for developing and executing complex systems. This makes it an ideal foundation for SOTA’s innovative toolkits.

Seamless Model Integration

Models developed with leading AI frameworks can be seamlessly integrated into the SOTA ecosystem. SOTA also enables your teams to reuse and optimize existing models across platforms, accelerating deployment while minimizing costs. As you can see here, flexibility means major hardware players are compatible.

Open Beta Content

For this Beta release, we are offering:

Available Toolkits:

Deep Learning Toolkit for LabVIEW: A comprehensive framework for loading, editing, training, and deploying neural networks directly within LabVIEW.
Computer Vision Toolkit for LabVIEW: Powerful tools for real-time image processing, object detection, and automated inspection.
Accelerator Toolkit for LabVIEW: Accelerates data processing using GPUs for faster computations and improved throughput.

Available Tools:

File Importator and Generator: Import and convert deep learning models (Keras, TensorFlow, PyTorch) with visualization and LabVIEW integration features.
Netron for LabVIEW: A viewer for analyzing neural network architectures.