cuda.tile.rsqrt

cuda.tile.rsqrt(x, /, *, flush_to_zero=False)

Perform rsqrt on a tile.

Parameters:

• x (Tile)

• flush_to_zero (const bool) – If True, flushes subnormal inputs and results to sign-preserving zero, default is False.

Return type:

Tile

Examples

SnippetComplete Example

tx = ct.full((4,), 4.0, dtype=ct.float32)
print(f"{ct.rsqrt(tx):.1f}")

Copy

import cuda.tile as ct
import torch

@ct.kernel
def kernel():
    tx = ct.full((4,), 4.0, dtype=ct.float32)
    print(f"{ct.rsqrt(tx):.1f}")


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

Output

[0.5, 0.5, 0.5, 0.5]