Graiphic extends ONNX from static inference to a unified graph that drives AI, logic, and hardware. SOTA, our ONNX-native LabVIEW IDE, makes this accessible across research, industry, and academia, with portable execution via ONNX Runtime.
Graiphic announces GO HW and SOTA, turning ONNX into a living execution plan for real systems. SOTA lets teams design and even train ONNX graphs visually in LabVIEW, choose device profiles, and run optimized sessions across CPU, GPU, FPGA, and SoC through ONNX Runtime Execution Providers. Workflows become reproducible and portable.
GO stands for graph orchestration. It brings control flow and scheduling into the graph. GO HW adds hardware-aware nodes such as DMA, GPIO, ADC, DAC and timers, enabling zero-copy paths, deterministic timing, and energy-per-inference measurement. Energy becomes a first-class metric with bench-tested evidence on Raspberry Pi, Jetson, Zynq, and industrial PCs.
Graiphic also bridges IR last with ONNX Runtime and IR first with MLIR, so European teams can pick portability or compiler-grade specialization under the same cockpit. This advances a sovereign, trustworthy, and energy-efficient AI stack for Europe.
Read more in the GO HW whitepaper.