Tensor (dpnp.tensor)

dpnp.tensor provides a reference implementation of the Python Array API specification. The implementation uses data-parallel algorithms suitable for execution on accelerators, such as GPUs.

dpnp.tensor is written using C++ and SYCL and oneAPI extensions implemented in Intel(R) oneAPI DPC++ compiler.

This module contains:

• Array object usm_ndarray

• Accumulation functions

• Array creation functions

• Array manipulation functions

• Data type functions

• Data types

• Elementwise functions

• Indexing functions

• Introspection functions

• Linear algebra functions

• Searching functions

• Set functions

• Sorting functions

• Statistical functions

• Utility functions

• Printing functions

• Constants