warp.force_load#

warp.force_load(
device=None,
modules=None,
block_dim=None,
max_workers=None,
)[source]#

Force user-defined kernels to be compiled and loaded (low-level API).

This is a lower-level function that accepts an explicit list of Warp Module objects. For most use cases, prefer load_module(), which provides a more convenient interface for loading modules by name or Python module reference.

Use this function when you:

  • Need to load multiple specific modules at once

  • Already have Module objects to work with

  • Want to load all modules containing Warp code (by passing modules=None)

Parameters:

  • device (Device | str | list[Device] | list[str] | None) – The device or list of devices to load the modules on. If None, load on all devices.

  • modules (list[Module] | None) – List of Warp Module objects to load. If None, load all imported modules that contain Warp code.

  • block_dim (int | None) – The number of threads per block (always 1 for "cpu" devices).

  • max_workers (int | None) – The maximum number of parallel threads to use for loading modules. 0 means serial loading. If None, `warp.config.load_module_max_workers determines the default.