| 摘要 |
Performance of a set of video processing devices in a flexible architecture can be predicted by analyzing and storing information about the performance of individual video processing devices. A combination of operations to be performed is associated with a set of video processing devices. Performance characteristics of individual video processing devices in the set are accessed. For each frame, a total performance characteristic of the set of video processing devices is determined by combining the performance characteristics of the individual video processing devices according to a combination of operations performed by the set of video processing devices to produce data for each frame. The performance characteristic may include latency, processing time and/or memory bandwidth. The total performance characteristic may be compared to a threshold. According to this comparison, it can be determined whether the set of video processing devices can produce each frame in real time.
|
