Search papers, labs, and topics across Lattice.
1
0
3
0
Achieve near-optimal throughput for concurrent DNN training and inference on edge devices by intelligently time-slicing workloads and dynamically adjusting power modes, even with limited profiling.
