# tree_d_3d.kernels.fmm_m2l_host

*module*

The fmm_m2l_host kernel of the FMM on the CPU: Pyccel and NumPy versions with the same
arguments.

The M2L translations grouped by destination cell: each cell sums the translated
expansions of all its partners in registers and writes its local expansion once, with
no atomics. The GPU runs ``fmm_m2l`` over the mutual pairs instead.

[View source](https://github.com/max-models/tree-d-3d/blob/devel/src/tree_d_3d/kernels/fmm_m2l_host/__init__.py#L1-L1)

## tree_d_3d.kernels.fmm_m2l_host.kernel

*attribute* · *module attribute*

```python
kernel = KERNEL_PACKAGE.from_folder(__name__)
```

[View source](https://github.com/max-models/tree-d-3d/blob/devel/src/tree_d_3d/kernels/fmm_m2l_host/__init__.py#L11-L11)

**Modules**

- `tree_d_3d.kernels.fmm_m2l_host.fmm_m2l_host_numpy`
- `tree_d_3d.kernels.fmm_m2l_host.fmm_m2l_host_pyccel`
