EXPLORE CONNECTIONS
W4A16 Mixed-Precision Matrix Multiplication on Decoupled Architecture: Kernel Design and Memory Bottleneck Analysis for Ascend NPUs
Follow the relationships that help you find your next source.
Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.