Search papers, labs, and topics across Lattice.
1
0
13
Achieve 10x lower quantization errors on MCUs with 16-bit precision, matching or beating 8-bit inference speed and energy costs.
