# Rectangular Y-junction power split vs Tidy3D

<span class="validation-badge">validated · 2026-08-28</span>

<div class="case-meta"><span>forward</span><span>resolution 1</span><span>parity</span><span>p4</span><span>pic</span><span>y_junction</span><span>tidy3d_golden</span><span>high_yield</span></div>

This case compares a local FDTDX result with pinned reference data generated by Tidy3D. The originating Tidy3D example is [`YJunction.ipynb`](https://github.com/flexcompute/tidy3d-notebooks/blob/c37c785d52e9258c9d048a781524b8e8d7c758ca/YJunction.ipynb), pinned at commit `c37c785d52e9`.

```{figure} ../../_static/generated/pic_geometry_atlas.svg
:alt: Technical schematic associated with Rectangular Y-junction power split vs Tidy3D

Integrated-photonics geometries covered by the benchmark suite.
```

## What is exercised

3D strip Y at dl≈40nm (~0.2M voxels). Compare mode-power T1,T2,Tsum to Tidy3D. Rectangular junction approx. Asymmetry or low Tsum = geometry/mode bug to fix.

This case runs at its declared full benchmark resolution (resolution factor 1.0).



## Recorded result

| Metric | Observed / error | Reference / limit | Error | Effective budget | Result |
|---|---:|---:|---:|---:|---:|
| `T1` | 0.202924 | 0.270472 | 0.0675476 | 0.147618 | <span class="metric-pass">PASS</span> |
| `T2` | 0.261742 | 0.270474 | 0.00873133 | 0.147618 | <span class="metric-pass">PASS</span> |
| `Tsum` | 0.464667 | 0.540946 | 0.0762789 | 0.208189 | <span class="metric-pass">PASS</span> |

The table is rendered from `progress.json`; it is not a hand-written success claim. For metrics that report an error directly, the “observed” column repeats that error and the reference column is the acceptance threshold.

## Reproduce

```bash
uv run fdtdx-bench run --case parity_y_junction
```

Implementation and provenance: `benchmarks/cases/parity_y_junction/case.yaml`, `benchmarks/cases/parity_y_junction/run.py`, `benchmarks/goldens/parity_y_junction.npz`.

[Return to the example atlas](../index.md)