AI RESEARCH
VTC: DNN Compilation with Virtual Tensors for Data Movement Elimination
arXiv CS.LG
•
ArXi:2604.09558v1 Announce Type: cross With the widening gap between compute and memory operation latencies, data movement optimizations have become increasingly important for DNN compilation. Current optimizations such as layout transformations and operator fusion only target a subset of tensor operators and consequently miss important opportunities for reducing data movement in contemporary DNN workloads, including large language models.