Skip to content
Snippets Groups Projects
user avatar
thorstenhater authored
Fuse kernels `gather` and `vec_minus` into a single kernel `set_dt_impl` for a small
performance improvement.

Here is the effect on the busyring benchmark (swapped pas -> hh) with 8192 cells on a 
V100 GPU (time for `model-run` in seconds).

```
|----------+-------|
| Baseline | After |
|----------+-------|
|    2.318 | 2.314 |
|    2.335 | 2.307 |
|    2.345 | 2.315 |
|    2.333 | 2.306 |
|    2.331 | 2.320 |
|----------+-------|
|    2.318 | 2.306 |
|----------+-------|
```
5f9f2a5a

CI status Build Status Gitpod Ready-to-Code

Arbor Library

Arbor is a library for implementing performance portable network simulations of multi-compartment neuron models.

An installation guide and library documentation are available online at Read the Docs.

Submit a ticket if you have any questions or want help.

Citing Arbor

The Arbor software can be cited via Zenodo: DOI.

Previous versions of Arbor can be cited specifically:

  • Version 0.2: DOI
  • Version 0.1: DOI

The following BibTeX entry can be used to cite Arbor:

@INPROCEEDINGS{
    paper:arbor2019,
    author={N. A. {Akar} and B. {Cumming} and V. {Karakasis} and A. {Küsters} and W. {Klijn} and A. {Peyser} and S. {Yates}},
    booktitle={2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP)},
    title={{Arbor --- A Morphologically-Detailed Neural Network Simulation Library for Contemporary High-Performance Computing Architectures}},
    year={2019}, month={feb}, volume={}, number={},
    pages={274--282},
    doi={10.1109/EMPDP.2019.8671560},
    ISSN={2377-5750}}

Alternative citation formats for the paper can be downloaded here, and a preprint is available at arXiv.