Mapping vision algorithms to parallel architectures
- Electrical Engineering and Computer Science Dept., Univ. of Michigan, Ann Arbor, MI (US)
Because parallel architectures can vary widely, the problem of mapping a parallel algorithm onto a given parallel machine is generally much harder than the problem of mapping a serial algorithm onto a serial machine. This paper examines some of the problems encountered, emphasizing mappings of vision algorithms onto mesh, hypercube, mesh-of-trees, pyramid, and Parallel Random Access Machines (PRAMs) having many simple processors, each with a small amount of memory. Approaches that have been suggested include simulating one architecture on another, designing ideal algorithms for ideal architectures and simulating the ideal architectures, and using general data movement operations. Each of these is shown to occasionally produce unacceptably inefficient implementations. It appears that as long as PRAMs cannot achieve the desired cost and performance goals, programmers must contend with carefully designing algorithms for specific architectures.
- OSTI ID:
- 6525907
- Journal Information:
- Proc. IEEE; (United States), Journal Name: Proc. IEEE; (United States) Vol. 76:8; ISSN IEEPA
- Country of Publication:
- United States
- Language:
- English
Similar Records
Parallel architectures and parallel algorithms for integrated vision systems
Parallel architectures and parallel algorithms for integrated vision systems