simple_nn.a

# Host mem: N x 8 input matrix
# Weight mem: 8 x 8 weight matrix
RHM 0, 0, 8   # read from host mem addr 0, to UB addr 0, for length N = 4
RW  0         # read weights from dram addr 0 to FIFO
MMC 0, 0, 8   # Do MM on UB addr 0, to accumulator addr 0, for length 4
ACT.Q 0, 0, 8 # Do ACT ReLU on accumulator addr 0, to UB addr 4, for length 4
SYNC          
WHM 0, 0, 8   # write result from UB addr 4, to host mem addr 4, for length 4
HLT