BRCloud Research Project

It is my pleasure to announce that our work is getting pursued in the new OmpCluster project (more information here). For this project, we have developed a new runtime implementation relying on MPI. The new runtime offers largely improved performance and task-based parallelization.

Read More

OmpCloud-based Ray-Tracer presented at WAMCA

We presented the implementation of our OmpCloud-based Raytracing Render Engine at the 9th Workshop on Applications for Multi-Core Architectures which took place during the SBAC-PAD conference at Lyon. The paper is available in the publications page.

Read More

New publications about OmpCloud

One of our paper presenting OmpCloud and the parallelization of a scientific application has been published in the ACM Transactions on Architecture and Code Optimization (TACO).

Read More

New Release v0.3.0

It is my pleasure to announce that OmpCloud v0.3.0 is now available. Get it here! This new version is adding support for older versions of Ubuntu (at least 14.04) and RPM-based distributions (experimental), and also adding support to loop scheduling model of OpenMP (static/dynamic, chuncksize).

Read More

New Release v0.2.0

It is my pleasure to announce that OmpCloud v0.2.0 is now available. Get it here! This new version is largely improving the usability of OmpCloud so do not be afraid to try it! Indeed, this release is the result of the feedback we got from the students which experimented OmpCloud during the course of parallel programming at UNICAMP. Please notice that the documentation available in the wiki has also been improved.

Read More

Talk at OpenMPCon 2017

Our abstract entitled “OmpCloud: Bridging the Gap between OpenMP and Cloud Computing” has been accepted for the OpenMP Developers Conference (OpenMPCon 2017) that will take place at Stony Brook University in New York, USA during September 18-20, 2017. People interested by our work can come there to see our talk!

Read More

OmpCloud Releases

From now, releases of OmpCloud will be made at regular time to allow the use of cloud offloading techniques without requiring to build Clang/LLVM by hand. However, we will propose binaries for linux-amd64 only.

Read More

Paper accepted to ICPP 2017

Our paper “The Cloud as an OpenMP Offloading Device” has been accepted to the 46th International Conference on Parallel Processing (ICPP-2017) that will take place in Bristol, UK during August 14-17, 2017.

Read More