cuda.tile.cat#

cuda.tile.cat(tiles, /, axis)#

Concatenates two tiles along the axis.

Parameters:

  • tiles (tuple) – a pair of tiles to concatenate.

  • axis (const int) – axis to concatenate the tiles.

Return type:

Tile

Notes

Due to power-of-two assumption on all tile shapes, the two input tiles must have the same shape.

Examples

tx = ct.full((2, 2), 3, dtype=ct.int32)
ty = ct.full((2, 2), 7, dtype=ct.int32)
print(ct.cat((tx, ty), 0))
print(ct.cat((tx, ty), 1))

import cuda.tile as ct
import torch

@ct.kernel
def kernel():
    tx = ct.full((2, 2), 3, dtype=ct.int32)
    ty = ct.full((2, 2), 7, dtype=ct.int32)
    print(ct.cat((tx, ty), 0))
    print(ct.cat((tx, ty), 1))


torch.cuda.init()
ct.launch(torch.cuda.current_stream(), (1,), kernel, ())
torch.cuda.synchronize()

Output

[[3, 3], [3, 3], [7, 7], [7, 7]]
[[3, 3, 7, 7], [3, 3, 7, 7]]