Optimization Models for Three On-Chip Network Problems
Nilay Vaish, Michael C. Ferris, David A. Wood · ACM Transactions on Architecture and Code Optimization · 2016
We model three on-chip network design problems—memory controller placement, resource allocation in heterogeneous on-chip networks, and their combination—as mathematical optimization problems. We model the first two problems as mixed integer linear programs. We model the third problem as a mixed integer nonlinear program, which we then linearize exactly. Sophisticated optimization algorithms enable solutions to be obtained much more efficiently. Detailed simulations using synthetic traffic and benchmark applications validate that our designs provide better performance than solutions proposed previously. Our work provides further evidence toward suitability of optimization models in searching/pruning architectural design space.
Cited by 4 works. doi:10.1145/2943781 Open-access PDF
Authors
References
- Convex Optimization
- Principles and Practices of Interconnection Networks
- The gem5 simulator
- Network Design and Transportation Planning: Models and Algorithms
- SPEC CPU2006 benchmark descriptions
- Lectures on Stochastic Programming
- Symbiotic jobscheduling for a simultaneous mutlithreading processor
- Symbiotic jobscheduling for a simultaneous multithreaded processor
- GARNET: A detailed on-chip network model inside a full-system simulator
- A polyhedral branch-and-cut approach to global optimization