Mojo function

ds_read_tr16_b64

ds_read_tr16_b64[dtype: DType, //](shared_ptr: LegacyUnsafePointer[Scalar[dtype], address_space=AddressSpace.SHARED, mut=mut, origin=origin]) -> SIMD[dtype, 4]

Reads a 64-bit LDS transpose block using TR16 layout and returns SIMD[dtype, 4] of 16-bit types.

Notes:

Only supported on AMD GPUs.
Maps directly to llvm.amdgcn.ds.read.tr16.b64 intrinsic.
Result width is fixed to 4 elements of dtype.

Parameters:

dtype (DType): Data type of the elements (must be 16-bit type).

Args:

shared_ptr (LegacyUnsafePointer): Pointer to the LDS transpose block.

Returns:

SIMD: SIMD[dtype, 4] of 16-bit types.