Abstract Available Parallelism in Video Applications
Abstract:
lelism has focused on general-purpose applications such as the SPEC benchmarks. Many quantitative experiments have been perjornted over the years measuring the impact of different execution models and optimization techniques on these applications. Recently, however, researchers have been developing various ILP architectures for media processors in order to exploit parallelism in audio, video, and graphics applications. It has been assumed that these applications contain far more potential parallelism than general-plrrpose code, but there have been ‘few attempts to quantify the available parallelism. In this paper, we present a linear complexity global scheduling algorithm that can process vev long traces up to I billion operations. Therefore, traces of video applications such as MPEGI, MPEG2, MPEG4 and H.263 encoders and decoders can be analyzed. Using an idealized esecution model, speedups of over 1000 have been found in some applications. The experiment shows that eliminating currently identifiable bottlenecks can allow the exploitation of huge anlounts of ILP in audio and video applications. 1
Citations
| 205 | Limits of control flow on parallelism – Lam, Wilson - 1992 |
| 182 | Available instruction-level parallelism for superscalar and superpipelined machines – Jouppi, Wall - 1989 |
| 72 | Measuring the Parallelism Available for Very Long Instruction Word Architectures – Nicolau, Fisher - 1984 |
| 1 | Three Vectors of Performance”, also http://www.mmx.com – Corp |
| 1 | Design Issue for a VcryLong-Instruction-Word VLSI Video – Dutta, Wolfe, et al. - 1996 |
| 1 | The Trimedia TM-I PC1 VLIW Mediaprocessor – Slavenburg - 1996 |
| 1 | Hardware/Software Interactions on Mpact”, pp. 20-26 – Kalapathy - 1997 |
