> For the complete documentation index, see [llms.txt](https://tu-clausthal.gitbook.io/good-research-practice-in-md-simulations-tuc-swz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://tu-clausthal.gitbook.io/good-research-practice-in-md-simulations-tuc-swz/infrastructures/hardware-options.md).

# Hardware Options

Choosing the right hardware for your MD simulations depends on the scale of your system, the length of the simulation, and the required turnaround time. The following options are available to SWZ researchers, ordered from most accessible to most powerful:

| Tier          | Resource                   | Cores                   | GPU            | Access              |
| ------------- | -------------------------- | ----------------------- | -------------- | ------------------- |
| Local         | Personal PC/Laptop         | 4-16                    | Optional       | Immediate           |
| Institutional | SWZ Rechencluster          | Up to 64/node (5 nodes) | T4, A100, H100 | Request via email   |
| National      | HLRN / NHR-Nord\@Göttingen | Thousands               | Yes (Grete)    | Project application |

## When to Use Each Tier

* **Personal PCs** -- code development, debugging input scripts, small test runs (<1000 atoms), visualization with OVITO
* **SWZ Cluster** -- production runs up to \~64 cores, GPU-accelerated simulations, medium-scale systems (up to \~10M atoms)
* **HLRN/NHR** -- large-scale production runs requiring hundreds to thousands of cores, very large systems (>10M atoms), or extensive parameter studies