# PINN Domain Decomposition

**URL:** <https://forums.developer.nvidia.com/t/pinn-domain-decomposition/263485>\
**Category:** Technical Support (PhysicsNeMo Only)\
**Tags:** modulus\
**Created:** [August 17, 2023, 6:31pm UTC](https://forums.developer.nvidia.com/t/pinn-domain-decomposition/263485 "2023-08-17T18:31:14Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![cpe.sk](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@cpe.sk](https://forums.developer.nvidia.com/u/cpe.sk)\
**Post date:** [August 17, 2023, 6:31pm UTC](https://forums.developer.nvidia.com/t/pinn-domain-decomposition/263485/1 "2023-08-17T18:31:15Z")

</div>

Hello everyone,

I would like to test my model on a large spatial domain. From what I read so far [1][2], it looks like a domain decomposition can enhance the prediction results. What is the best way to implement this in Modulus? Is it best to modify the conf files or create a new network model? Is it possible to modify any of the current network models to perform domain decomposition?

[1] Kharazmi, Ehsan, Zhongqiang Zhang, and George Em Karniadakis. “hp-VPINNs: Variational physics-informed neural networks with domain decomposition.” Computer Methods in Applied Mechanics and Engineering 374 (2021): 113547.

[2] N. Borrel-Jensen, S. Goswami, A. P. Engsig-Karup, G. E. Karniadakis, and C.-H. Jeong, “Sound propagation in realistic interactive 3D scenes with parameterized sources using deep neural operators,” pp. 1–25, 2023.

Thank you for your support!

---

<div class="post-metadata">

**Author:** ![tjain](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@tjain](https://forums.developer.nvidia.com/u/tjain)\
**Post date:** [December 8, 2023, 4:14pm UTC](https://forums.developer.nvidia.com/t/pinn-domain-decomposition/263485/2 "2023-12-08T16:14:50Z")

</div>

Looking for the same. There are some examples available here:

> **[GitHub - ehsankharazmi/hp-VPINNs: hp-VPINNs: variational physics-informed neural...](https://github.com/ehsankharazmi/hp-VPINNs)**
>
> hp-VPINNs: variational physics-informed neural network with domain decomposition is a general framework to solve differential equations

But haven’t found a way to integrate in modulus.