RAFI -- A Ray/Work Forwarding Infrastructure for Data Parallel Multi-Node/Multi-GPU Computing
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914614607872000 |
|---|---|
| author | Wald, Ingo Demirci, Serkan Sahistan, Alper Zellmann, Stefan Paris, Andrea Moran, Patrick Jaros, Milan von Landesberger, Tatiana Gudukbay, Ugur Pascucci, Valerio |
| author_facet | Wald, Ingo Demirci, Serkan Sahistan, Alper Zellmann, Stefan Paris, Andrea Moran, Patrick Jaros, Milan von Landesberger, Tatiana Gudukbay, Ugur Pascucci, Valerio |
| contents | We present RaFI, a CUDA and MPI based software framework that simplifies the task of building GPU-enabled data-parallel software where rays or similar work items need to migrate between different GPUs. RaFI provides a simple interface for CUDA kernels to forward such work items to other GPUs, while under the hood managing all the CUDA and MPI related work required to make this happen. We describe RaFI's motivation and implementation, and show its potential in several example applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_30294 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | RAFI -- A Ray/Work Forwarding Infrastructure for Data Parallel Multi-Node/Multi-GPU Computing Wald, Ingo Demirci, Serkan Sahistan, Alper Zellmann, Stefan Paris, Andrea Moran, Patrick Jaros, Milan von Landesberger, Tatiana Gudukbay, Ugur Pascucci, Valerio Distributed, Parallel, and Cluster Computing Graphics We present RaFI, a CUDA and MPI based software framework that simplifies the task of building GPU-enabled data-parallel software where rays or similar work items need to migrate between different GPUs. RaFI provides a simple interface for CUDA kernels to forward such work items to other GPUs, while under the hood managing all the CUDA and MPI related work required to make this happen. We describe RaFI's motivation and implementation, and show its potential in several example applications. |
| title | RAFI -- A Ray/Work Forwarding Infrastructure for Data Parallel Multi-Node/Multi-GPU Computing |
| topic | Distributed, Parallel, and Cluster Computing Graphics |
| url | https://arxiv.org/abs/2605.30294 |