Mapping tensor products onto VLSI networks with reduced I/O | IEEE Conference Publication | IEEE Xplore