The Lattice Semiconductor Compact CNN Accelerator IP Core is a calculation engine for Deep Neural Network with fix point weights or binarized weights. It calculates full layers of neural network including convolution layer, pooling layer, batch normalization layer and full connect layer by executing sequence code with weight values which is generated by the Lattice Neural Network Compiler tool. Engine is optimized for convolutional neural network, which is why it can be used for vision-based application such as classification or face detection and key phrase detection. The engine is not a simple accelerator which requires an extra processor; it can do all required calculations by itself.
iCE40UP
2.1.0 | Added configuration attribute: No. of 1x1 Conv. Engines. Added o_gpo output signal. |
2.0.0 | Added configuration attribute: Byte Mode, Byte Shift. Updated possible values of Memory Type attribute. |
1.1.0 | This IP is re-encypted for Radiant 1.1. Not compatible with Radiant 1.0. |
1.0.0 | Initial release. |