Abstract
An optimal processor allocation technique for parallelizing compositing back-end for sort-last rendering systems based on a BSP-tree execution model is proposed. This allocation technique can be readily adapted to optimize other system parameters, such as frame rate and latency. Other techniques are also introduced to optimize sequential operations of individual compositing processes, as well as to reduce communication overheads and effectively pipeline compositing operations. For purposes of illustration, all these techniques are implemented on a MPP machine.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings of SPIE - The International Society for Optical Engineering |
| Publisher | Society of Photo-Optical Instrumentation Engineers |
| Pages | 182-192 |
| Number of pages | 11 |
| Volume | 3643 |
| State | Published - 1999 |
| Event | Proceedings of the 1999 Visual Data Exploration and Analysis VI - San Jose, CA, USA Duration: Jan 27 1999 → Jan 28 1999 |
Other
| Other | Proceedings of the 1999 Visual Data Exploration and Analysis VI |
|---|---|
| City | San Jose, CA, USA |
| Period | 1/27/99 → 1/28/99 |
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Condensed Matter Physics