ifdp · June 30, 1995

Block Distributed Methods for Solving Multi-Country Econometric Models

Abstract

This paper examines variations on a baseline Fair-Taylor algorithm used to solve multi-country, rational expectations models. One notable feature of these variations is the ability to exploit small-scale distributed processing using a network of workstations or PCs. Using four processors to solve MX-4 (152 endogenous variables), the largest speedup factor relative to Fair-Taylor is 59; for RE-7 (978 endogenous variables) the maximum speedup factor is 12.

Abstract This paper examines variations on a baseline Fair-Taylor algorithm used to solve multi-country, rational expectations models. One notable feature of these variations is the ability to exploit small-scale distributed processing t.sing a network of workstations or pes. Using fouf processors to solve MX-4 (152 endogenous variables), the largest speedup factor relative to Fair-Taylor iB 59; for RE-7 (978 endogenous variables) the maximum speedup factor is 12.

Cite this document
APA
Jon Faust and Ralph Tryon (1995). Block Distributed Methods for Solving Multi-Country Econometric Models (IFDP 1995-516). Board of Governors of the Federal Reserve System, International Finance Discussion Papers. https://whenthefedspeaks.com/doc/ifdp_1995-516
BibTeX
@techreport{wtfs_ifdp_1995_516,
  author = {Jon Faust and Ralph Tryon},
  title = {Block Distributed Methods for Solving Multi-Country Econometric Models},
  type = {International Finance Discussion Papers},
  number = {1995-516},
  institution = {Board of Governors of the Federal Reserve System},
  year = {1995},
  url = {https://whenthefedspeaks.com/doc/ifdp_1995-516},
  abstract = {This paper examines variations on a baseline Fair-Taylor algorithm used to solve multi-country, rational expectations models. One notable feature of these variations is the ability to exploit small-scale distributed processing using a network of workstations or PCs. Using four processors to solve MX-4 (152 endogenous variables), the largest speedup factor relative to Fair-Taylor is 59; for RE-7 (978 endogenous variables) the maximum speedup factor is 12.},
}