Mojo function

global_load_dword

global_load_dword[dtype: DType, width: Int](sgpr_address: SIMD[DType.uint32, 2], vgpr_offset: UInt32) -> SIMD[dtype, width]

Loads 8 or 16 bytes from global memory to a SIMD vector. Each load is asynchronous and needs to be waited on with wait_vmcount.

Parameters:

• ​dtype (DType): The data type of global memory.
• ​width (Int): The width of the SIMD vector to load.

Args:

• ​sgpr_address (SIMD): Two consecutive SGPRs that contain the address of the global memory ptr.
• ​vgpr_offset (UInt32): The offset of the address that this thread will load from.

Returns:

SIMD

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit