The purpose of OpenSWR is to provide a high performance, highly scalable OpenGL compatible software rasterizer that allows use of unmodified visualization software. This allows working with datasets where GPU hardware isn't available or is limiting. OpenSWR is completely CPU-based, and runs on anything from laptops, to workstations, to compute nodes in HPC systems.
Click either image for more information
OpenSWR internally builds on top of LLVM, and fully utilizes modern instruction sets like Intel® and Intel® Advanced Vector Extensions (AVX and AVX2) to achieve high rendering performance. The charts below illustrate the compelling advantage of OpenSWR over Mesa llvmpipe in a real application.
Click for more information
OpenSWR is now fully integrated into Mesa and provides an SWR renderer that supports much of the OpenGL 3.3 Core and OpenGL 3.0 Compatibility contexts. Standard Mesa environment variables provide the ability to run-time switch between OpenSWR and llvmpipe software renderers.
For more detail, please see the "OpenSWR Overview" presented at the Intel® HPC Developers Conference at SC15.
How to build Mesa with OpenSWR:
On Linux, OpenSWR is built using the standard autoconf package as described here, with the addition of "swr" to the list of gallium drivers.
- July 8, 2016
Mesa 12.0 has been released and contains OpenSWR as an optional driver. Development work on OpenSWR is now done on the mesa git master repository.
- Oct 19, 2015
Our initial public version of our Mesa+SWR integration work is now available from https://github.com/OpenSWR/openswr-mesa.
- Aug 9, 2015
Since the release of early alpha, we have been busy integrating our OpenSWR core into the Mesa project. This enables us to take advantage of a very mature, very feature complete driver stack that would have been very difficult to develop on our own.
The result is that OpenSWR now has far greater functionality than the OpenGL 1.4 features available in the first alpha release.
- Dec 12, 2014
ALPHA release. This version is the first to be released to the public. Please bear with us if there are build or functionality issues. The major applications used for testing were ParaView and VisIt - if you are trying another program you may encounter missing features.
This version is coming out as we are working on some major cleanups to the code. In order to keep our commitment to release this to the community and to provide a well-tested version, we are releasing the code that was used for the SC14 demonstrations. The next major release will contain the cleanups.
The source code can be downloaded from here:
- Dec 8, 2014
- The slides for our OpenSWR-related talks at SC14 are now available:
- Nov 13, 2014
There is an upcoming talk on OpenSWR at Supercomputing 2014:
- At the "Visualization" track of the Intel® HPC Developer Conference (Nov 16, 2014)