# tf.distribute.experimental.CommunicationImplementation
|---|
| [ View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.14.0/tensorflow/python/distribute/collective_util.py#L26-L44) |
Cross device communication implementation.
#### View aliases
**Main aliases**
[`tf.distribute.experimental.CollectiveCommunication`](https://www.tensorflow.org/api_docs/python/tf/distribute/experimental/CommunicationImplementation)
**Compat aliases for migration**
See
[Migration guide](https://www.tensorflow.org/guide/migrate) for
more details.
[`tf.compat.v1.distribute.experimental.CollectiveCommunication`](https://www.tensorflow.org/api_docs/python/tf/distribute/experimental/CommunicationImplementation), [`tf.compat.v1.distribute.experimental.CommunicationImplementation`](https://www.tensorflow.org/api_docs/python/tf/distribute/experimental/CommunicationImplementation)
> [!WARNING]
> **Warning:** The alias [`tf.distribute.experimental.CollectiveCommunication`](https://www.tensorflow.org/versions/r2.14/api_docs/python/tf/distribute/experimental/CommunicationImplementation) is deprecated and will be removed in a future version. Use [`tf.distribute.experimental.CommunicationImplementation`](https://www.tensorflow.org/versions/r2.14/api_docs/python/tf/distribute/experimental/CommunicationImplementation) instead.
- `AUTO`: Automatically chosen by Tensorflow.
- `RING`: TensorFlow's ring algorithms for all-reduce and all-gather.
- `NCCL`: NVIDIA®'s NCCL library. This is now only used for all-reduce on GPUs; all-reduce on CPU, all-gather and broadcast fallbacks to RING.
| ## Class Variables ||
|---|---|
| AUTO | `` |
| NCCL | `` |
| RING | `` |