Support unpadded shapes in matmul1d w/ gather_in0

[avoraTT]

Problem

Currently, Matmul1D with gather_in0=True does not handle shapes (K, N) that do not divide by the number of cores in the ring. In the current use case in the Llama models, the activations and weights need to be padded in order to use the matmul. However, this results in significant overhead caused by padding/slicing. Therefore, this matmul must support shapes that do not divide with the number of cores, and handle the padding implicitly.