warp.config.enable_mempools_at_init

warp.config.enable_mempools_at_init : bool = True

Enable CUDA memory pools during device initialization when supported.

Set WARP_ENABLE_MEMPOOLS_AT_INIT to 0, false, or no before importing Warp to disable automatic memory pool initialization (e.g. when sharing the process with a framework that owns its own GPU allocator).