Comparisons
===========

MOA is not unique in its attempt to be a high level tensor
compiler. Thus we find it important to contrast it with others
approaches and show the landscape of available compilers.

 - `MLIR (Google) <https://github.com/tensorflow/mlir>`_
 - `PlaidML (Intel) <https://github.com/plaidml/plaidml>`_
 - `TensorComprehensions (Facebook) <https://github.com/facebookresearch/TensorComprehensions>`_
 - `ngraph (Intel) <https://github.com/NervanaSystems/ngraph>`_
 - `taco <https://github.com/tensor-compiler/taco>`_
 - `tvm <https://github.com/dmlc/tvm>`_
 - `xla (Google) <https://github.com/tensorflow/tensorflow/tree/master/tensorflow/python/compiler/xla>`_