CompiledCudaComplexEvaluator

Symbolica documentation for getting started, symbolic expressions, numerical evaluation, pattern matching, and APIs in Python and Rust.

CompiledCudaComplexEvaluator

CompiledCudaComplexEvaluator()

A compiled evaluator of an expression that uses CUDA for GPU acceleration.

Methods

Name Description
evaluate Evaluate the expression for multiple inputs and return the result.
load Load a compiled library, previously generated with Evaluator.compile()

evaluate

CompiledCudaComplexEvaluator.evaluate(inputs: npt.ArrayLike) -> npt.NDArray[np.complex128]

Evaluate the expression for multiple inputs and return the result.

Parameters

  • inputs (npt.ArrayLike) The input values or batches to evaluate.

load

CompiledCudaComplexEvaluator.load(
    filename: str,
    function_name: str,
    number_of_evaluations: int,
    block_size: int = 512,
) -> CompiledCudaComplexEvaluator

Load a compiled library, previously generated with Evaluator.compile(). Input and output dimensions are read from the library. Libraries generated before Symbolica 3.0 must be re-exported and recompiled.

Parameters

  • filename (str) The file path to load from or save to.
  • function_name (str) The exported symbol name of the compiled entry point.
  • number_of_evaluations (int) The number of evaluations to batch per CUDA kernel launch.
  • block_size (int) The CUDA thread block size used by the compiled kernel.