Matrix Multiplication Based on Scalable Macro-Pipelined FPGA Accelerator Architecture | IEEE Conference Publication | IEEE Xplore