Abstract
In this paper, we propose new architecture that has some PU(Processing Unit) arrays of Nearest Neighbor Mesh(NNM), Using some NNM PU arrays, it is faster than one PU array when calculate the rolling mapping for the mesh with various densities.
Original language | English |
---|---|
Pages (from-to) | 233-245 |
Number of pages | 13 |
Journal | Advances in High Performance Computing |
Volume | 6 |
Publication status | Published - 2000 Dec 1 |
Event | Sixth International Conference on Applications of High-Performance Computers in Engineering - Maui, HA, United States Duration: 2000 Jan 26 → 2000 Jan 28 |
ASJC Scopus subject areas
- Computer Science(all)
- Engineering(all)