qp.backline.css_bp_decoder¶
- css_bp_decoder(Hx, Hz, *, postprocess='osd', num_iters=10, prob=0.1, **build_options)[source]¶
Compile a CSS code’s Tanner graph into a coprocessor decode function for use with
backline.Accepts the X- and Z-type parity-check matrices of a CSS code and compiles a decoder down to a shared library that can be used as a
CoprocessorFunction.Warning
Backlineis experimental and only usable through the Catalyst compiler.- Parameters:
Hx (ArrayLike) – X parity-check matrix.
Hz (ArrayLike) – Z parity-check matrix.
- Keyword Arguments:
postprocess (str) – Postprocessing step applied after belief propagation. Use
"hard"for hard-decision output or"osd"for ordered-statistics decoding.num_iters (int) – Number of belief-propagation iterations.
prob (float) – Uniform prior error probability across qubits.
platform (str) – Required Triton platform string of the form
"backend:arch:warp_size". For example,"hip:gfx942:64"or"cuda:80:32".grid (tuple[int, int, int]) – Triton kernel launch grid dimensions. Defaults to
(1, 1, 1).num_warps (int) – Triton kernel launch warp count. Defaults to
1.num_stages (int) – Triton pipeline stage count. Defaults to
1.compiler (str) – Optional compiler executable override. Defaults to
"".cflags (tuple[str, ...]) – Extra compiler flags. Defaults to
().
- Returns:
- The compiled decode function, ready to pass as
- Return type:
- Raises:
ImportError – If Triton decoder support is unavailable.
ValueError – If the decoder options or parity-check matrices are invalid.
See also
Example
>>> import numpy as np >>> import pennylane as qp >>> Hz = Hx = np.array([ ... [1, 0, 1, 0, 1, 0, 1], ... [0, 1, 1, 0, 0, 1, 1], ... [0, 0, 0, 1, 1, 1, 1], ... ]) >>> decoder = qp.backline.css_bp_decoder( ... Hx, ... Hz, ... postprocess="hard", ... num_iters=5, ... platform="hip:gfx942:64", ... )