Search papers, labs, and topics across Lattice.
Shahid Bahonar University, Kerman, Iran
2
0
4
Achieve 330x energy reduction in spiking neural networks by adaptively exiting computation based on input complexity using reinforcement learning.
Forget exhaustive search: HAWX achieves up to 3 million times speedup in DNN approximation by intelligently integrating heterogeneous approximate computing blocks, guided by multi-level sensitivity analysis.
