TF3810

TF3810 | TC3 Neural Network Inference Engine

The TF3810 TwinCAT 3 function is a high-performance execution module (inference machine) for trained neural networks. The neural networks are trained in established frameworks such as PyTorch, TensorFlow or MATLAB®. The information of the learned network is loaded to the inference machine as a description file. The execution module can be called in TwinCAT 3 from the PLC, from C++ and via a cyclic task. The neural network is executed directly on the machine controller in the TwinCAT real-time cycle.

Technical data TF3810
Required TC1000
Target system Windows 7/10
Ordering information
TF3810-0v40 TC3 Neural Network Inference Engine, platform 40 (Performance)
TF3810-0v50 TC3 Neural Network Inference Engine, platform 50 (Performance Plus)
TF3810-0v60 TC3 Neural Network Inference Engine, platform 60 (Mid Performance)
TF3810-0v70 TC3 Neural Network Inference Engine, platform 70 (High Performance)
TF3810-0v80 TC3 Neural Network Inference Engine, platform 80 (Very High Performance)
TF3810-0v81 TC3 Neural Network Inference Engine, platform 81 (Many-core 5…8 Cores)
TF3810-0v82 TC3 Neural Network Inference Engine, platform 82 (Many-core 9…16 Cores)
TF3810-0v83 TC3 Neural Network Inference Engine, platform 83 (Many-core 17…32 Cores)
TF3810-0v84 TC3 Neural Network Inference Engine, platform 84 (Many-core 33…64 Cores)
TF3810-0v90 TC3 Neural Network Inference Engine, platform 90 (Other)
TF3810-0v91 TC3 Neural Network Inference Engine, platform 91 (Other 5…8 Cores)
TF3810-0v92 TC3 Neural Network Inference Engine, platform 92 (Other 9…16 Cores)
TF3810-0v93 TC3 Neural Network Inference Engine, platform 93 (Other 17…32 Cores)
TF3810-0v94 TC3 Neural Network Inference Engine, platform 94 (Other 33…64 Cores)
Product announcement
estimated market release 2nd quarter 2020