AMATEROUS: A PARALLEL WIRE ROUTER WITHOUT DETAILED/GLOBAL FEEDBACK

This paper proposes a new parallel wire routing algorithm named Amaterous. As many successful parallel wire routers, Amaterous has global and detailed routers, but the feedback from the detailed to the global is removed so that both routers are parallelized e ciently and independently. For this feedback removal, Amaterous has an additional routing phase named capacity path (c-path) router that draws all the possible paths in each rectangular region of each routing layer prior to the global routing. Since the global router picks c-paths to form the global path of a net, it is assured that the global path is always converted successfully to its detailed counterpart which the detailed router generates from the c-paths. Therefore the global router does not need to know the result of the detailed routing and thus the feedback from detailed to global is removed. We implemented Amaterous on a 16-node PC cluster and measured its performance with three MCM benchmarks. The result shows up to 12 fold speedup proving the e ciency of our algorithm. It is also shown that Amaterous is 1.2 to 3.6 times as fast as a state-of-art parallel router Amon2.

[1]  Prithviraj Banerjee,et al.  PHIGURE: a parallel hierarchical global router , 1990, 27th ACM/IEEE Design Automation Conference.

[2]  Jonathan Rose LocusRoute: a parallel global router for standard cells , 1988, 25th ACM/IEEE, Design Automation Conference.Proceedings 1988..

[3]  Jason Cong,et al.  An Efficient Multilayer MCM Router Based on Four-Via Routing , 1993, 30th ACM/IEEE Design Automation Conference.

[4]  Yoshizo Takahashi,et al.  Parallel automated wire-routing with a number of competing processors , 1990, ICS '90.

[5]  Nobuhiko Koike,et al.  PROTON: a parallel detailed router on an MIMD parallel machine , 1991, 1991 IEEE International Conference on Computer-Aided Design Digest of Technical Papers.

[6]  Kunle Olukotun,et al.  A Preliminary Investigation into Parallel Routing on a Hypercube Computer , 1987, 24th ACM/IEEE Design Automation Conference.

[7]  K. Taki,et al.  A parallel lookahead line search router with automatic ripup-and-reroute , 1993, 1993 European Conference on Design Automation with the European Event in ASIC Design.

[8]  Hiroshi Nakashima,et al.  Amon: a parallel slice algorithm for wire routing , 1995, ICS '95.

[9]  Hitoshi Kitazawa,et al.  A Parallel Adaptable Routing Algorithm and its Implementation on a Two-Dimensional Array Processor , 1987, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.