| | 5 | == Installation == |
| | 6 | |
| | 7 | Best method for installing Equalizer library is to compile it from sources. Currently installed version on 16K wall is the latest 1.2. Any required packages that are missing are reported during cmake configuration. It is best to keep build directories in the same path on each node of the system. |
| | 8 | |
| | 9 | == Configuration == |
| | 10 | |
| | 11 | Equalizer can work in numerous modes, including distributed rendering for single display, walls, caves, all in planar 2D or 3D. Everything is setup using proper configuration file. You can read more about these files in documentation and user's manual or look at sample configuration files in examples/configuration directory. Configs for 16K wall can be found on svn. To pass specific configuration file you have to use --eq-config switch when running application. |
| | 12 | |
| | 13 | == Differences with SAGE == |
| | 14 | |
| | 15 | The main difference between using Equalizer and SAGE is method of image acquisition. In SAGE the whole, uncompressed image is sent to tiles over the network, which requires very high bandwidth. Equalizer on the other hand uses network only to synchronize between nodes, everything is renedered on tiles. Therefore, when playing video or rendering model from file each node needs to have a copy of it located in the same path. Of course, we can use network storage, but then again high-speed network is required to send all the data to tiles. |
| | 16 | |
| | 17 | To sum up, SAGE's bottleneck is network bandwidth, Equalizer's bottleneck is GPU power and I/O speed. |
| | 18 | |
| | 19 | == Sample applications == |
| | 20 | |
| | 21 | Equalizer comes with few samples, like eVolve used to render |
