Generate GPU code

Enable code generation for GPUs

Since R2020b

Model Configuration Pane: Code Generation

Description

The Generate GPU code parameter specifies whether to use GPU Coder™ and generate CUDA^® code.

Dependencies

• This parameter requires a GPU Coder license.

• To enable this parameter, select C++ for the Language and grt.tlc or ert.tlc for System target file.

• Enabling Generate GPU code selects Support long long and Support long long parameters on the Hardware Implementation pane.

• For an GRT-based system target file, enabling Generate GPU code changes the following parameters on the Code Generation > Optimization pane:

  • Selects None for Leverage target hardware instruction set extensions.

  • Enables Use memset to initialize floats and doubles to 0.0.

• For an ERT-based system target file, enabling Generate GPU code changes the following parameters on the Code Generation > Optimization pane:

  • Disables Generate parallel for-loops (Embedded Coder).

  • Selects <empty> for Disable incompatible optimizations.

• For an ERT-based system target file, enabling Generate GPU code selects Terminate function required (Embedded Coder) parameter on the Code Generation > Interface pane.

Settings

off (default) | on

On

Enables code generation by using GPU Coder.

Off

Disables code generation by using GPU Coder.

Recommended Settings

Application	Setting
Debugging	No impact
Traceability	No impact
Efficiency	No impact
Safety precaution	No impact

Programmatic Use

Parameter: GenerateGPUCode

Type: character vector

Value: 'CUDA' | 'none'

Default: 'none'

Version History

Introduced in R2020b