Scalable-ABM Literature

Relevant literature for the Scalable ABM Paper

HPC Literature

  1. Speedup Versus Efficiency in Parallel Systems. Eager.
  2. Validity of the single processor approach to achieving large scale computing capabilities. Amdahl.
  3. Reevaluating Amdahl's law. Gustafson.
  4. Performance Properties of Large Scale Parallel Systems. Gupta.
  5. Analyzing Scalability of Parallel Algorithms and Architectures. Kumar.
  6. Isoefficiency: measuring the scalability of parallel algorithms and architectures. Grama.
  7. Scalable Load Balancing Techniques for Parallel Computers. Kumar.
  8. Evolutionary games and computer simulations. Huberman.
    • Shows the difference between sync and async.

Extra reading

  1. Petascale Computational Systems: Balanced CyberInfrastructure in a Data-Centric World. Bell.
  2. Some Essential Techniques for Developing Efficient Petascale Applications. Kale.
  3. The free lunch is over: A fundamental turn toward concurrency in software. Sutter.

- M.P. Armstrong, R.J. Marciano, International Journal of Geographical Information Systems 10 (6) (1996) 713–730. ?? === ABM Literature === - Parallel Individual-Based Modeling of Everglades Deer Ecology. Abbott. * Row decomp instead of block decomp * 32 procs - Parallel simulation of ecological structured communities: Computational needs, hardware capabilities, and nonlinear applications. Nichols. * scaling * load-imbalance * 256 procs - Parallel simulation of individual-based, physiologically structured population models. Ramachandramurthi. * Defines load-imbalance, and presents load-balancing algorithm. * look at different parallelization strategies * 32 procs - Agent-Based Computational Models And Generative Social Science. Epstein. * Define computational ABM to differentiate from statistical models * Bottom-up approach to study non-equilibrium systems and see macro scale effects * Agents have bounded rationality and we have bounded computation * Mentions sugarscape in example (vision increase) * Computational characteristics of 'hard social problems' - Adaptive Algorithms for the Dynamic Distribution and Parallel Execution of Agent-Based Models. Scheutz. * Define parallel ABM * formal definition - may be able to use ABM/GPU - A Framework for Megascale Agent Based Model Simulations on the GPU. Lysenko. * Explains ping/pong technique for updates (similar to red/black jacobi) * Collision map - approach to handling conflicting agent movement * Looks at handling death/reproduction - Sugarscape on Steroids: Simulating over a Million Agents at Iterative Rates. D'Souza. * first GPU implementation * defines some basic sugarscape rules * Collision map (same as above paper) using scatter * Good performance * 1 gpu === Simulation Literature === - Parallel implementation of the TRANSIMS micro-simulation. Nagel. * CA based simulation * domain decomp * adaptive loadbalancing * 1024 cpus - Geographic Information Science: The Grand Challenges. Goodchild. * Use this sub-goal in digital earth to justify this work β€œTo create simulations of geographic phenomena in a digital computer that are indistinguishable from their real counterparts (the modeling challenge; in effect a Turing test of GIS-based modeling).” - HPC and Geography. Openshaw. * Demonstrates need for HPC and Geography * Provides a few modeling examples as good use cases Return to index