About backends

What’s a backend?

In the nGraph Compiler stack, what we call a backend is responsible for function execution and value allocation. A backend can be used to carry out a programmed computation from a framework on a CPU or GPU; or it can be used with an Interpreter mode, which is primarily intended for testing, to analyze a program, or to help a framework developer customize targeted solutions. Experimental APIs to support current and future nGraph Backends are also available; see, for example, the section on PlaidML from nGraph.

Backend Current nGraph support Future nGraph support
Intel® Architecture Processors (CPUs) Yes Yes
Intel® Nervana™ Neural Network Processor™ (NNPs) Yes Yes
NVIDIA* CUDA (GPUs) Yes Some
AMD* GPUs Yes Some

How to use?

  1. Create a Backend; think of it as a compiler.
  2. A Backend can then produce an Executable by calling compile.
  3. A single iteration of the executable is executed by calling the call method on the Executable object.
../_images/execution-interface.png

The execution interface for nGraph

The nGraph execution API for Executable objects is a simple, five-method interface; each backend implements the following five functions:

  • The create_tensor() method allows the bridge to create tensor objects in host memory or an accelerator’s memory.
  • The write() and read() methods are used to transfer raw data into and out of tensors that reside in off-host memory.
  • The compile() method instructs the backend to prepare an nGraph function for later execution.
  • And, finally, the call() method is used to invoke an nGraph function against a particular set of tensors.