A bridging model for parallel computation
Communications of the ACM
Machine models and simulations
Handbook of theoretical computer science (vol. A)
General purpose parallel computing
Lectures on parallel computation
Journal of the ACM (JACM)
A universal interconnection pattern for parallel computers
Journal of the ACM (JACM)
Weak Parallel Machines: a New Class of Physically Feasible Parallel Machine Models
MFCS '92 Proceedings of the 17th International Symposium on Mathematical Foundations of Computer Science
On the Distributed Realization of Parallel Algorithms
SOFSEM '97 Proceedings of the 24th Seminar on Current Trends in Theory and Practice of Informatics: Theory and Practice of Informatics
Horizons of Parallel Computation
Horizons of Parallel Computation
Decomposable Bulk Synchronous Parallel Computers
SOFSEM '99 Proceedings of the 26th Conference on Current Trends in Theory and Practice of Informatics on Theory and Practice of Informatics
Hi-index | 0.00 |
We show how the speed of a BSP (Bulk Synchronous Parallel) computer depends on the parameters p, g, and l of the model. According to the values of parameters, BSP belongs among the first (sequential) or the second (parallel) class models or neither of these classes. The relation between BSP and the class of weak parallel machines is also examined. It turns out that BSP does not fit to the concept of weak (or pipelined) parallelism. Consequences of membership in different machine classes to the physical feasibility of BSP computers are discussed. The main conclusion is that BSP with parameters properly chosen qualifies itself as a practical model, but it is unable to exploit all the parallelism allowed by laws of physics.