high performance geo-simulations
DESCRIPTION
High Performance Geo-Simulations. Saulo Henrique Cabral Silva Tiago Garcia de Senna Carneiro* TerraLAB – Laboratory for Earh System Modeling and Simulation Computer Science Department Federal University of Ouro Preto GeoInfo 2013. Challenges. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/1.jpg)
High Performance Geo-SimulationsSaulo Henrique Cabral Silva
Tiago Garcia de Senna Carneiro*
TerraLAB – Laboratory for Earh System Modeling and SimulationComputer Science Department
Federal University of Ouro Preto
GeoInfo 2013
![Page 2: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/2.jpg)
Challenges• Modelers have few programming expertise
– Problem partition into several processes?– Processes communication?– Processes synchronization?
• What is the design goal: Scalability or Performance?– Global scheduling scalabitity– Local scheduling high performance
• How to parallelize/distribute workload?– Model architecture can be defined by:
• Application domain area (the phenomenon being modeled)• Chosen modeling paradigm: CA, multiagente, DEVS, etc
– Which load balancing strategy to use?• Agents clusters• Dynamic and adaptative
• How to verify models with no influence of concurrency control rules?
![Page 3: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/3.jpg)
function execAgentPred(i) predatores[i]:execute() end
function execAgentPrey(i) preys[i]:execute()
end
for i = 1, #predatores do --HPA PARALLEL
execAgentPred(i) end --HPA JOINALL for i = 1, #preys do
--HPA PARALLEL execAgentPrey(i) end --HPA JOINALL
Annotations - API:
HPA PARALLEL HPA JOIN function_name HPA JOINALL HPA ACQUIRE resource_name HPA RELEASE resource_name
Flexibility through annotation: Prey-predator model
![Page 4: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/4.jpg)
Scalability through load balance: Bag of Tasks
![Page 5: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/5.jpg)
Discussion• TerraME HPA scales well when the modeler designs efficient
parallel solutions
– 90% of linear speedup with 8 processors (prey predator model)
– 35% of linear speedup with 8 processors (TROLL model)
![Page 6: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/6.jpg)
EXP1: Spatially-explicit prey predator model
• Species compete, evolve and disperse, in order to seek resources to support their struggle for their existence
• Each cell has a population of prey and a population of predators
• Model executes in two steps:– First, predators are simulated in parallel– Then, prey are simulated in parallel
![Page 7: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/7.jpg)
![Page 8: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/8.jpg)
![Page 9: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/9.jpg)
Distributed (Dmason) versus Parallel (TerraME)
![Page 10: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/10.jpg)
Agents Clusters versus Load Balancing
![Page 11: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/11.jpg)
EXP2: 3D tropical rain forest growth model - TROLL
• TROLL model simulates a 3D environment where:– Individual trees compete for light– Seed dispersal by wind– Trees interacts falling over others and forming gaps– …
• Adopted partitioning strategy: each layer (h, h-1, h-2, …) is simulated in parallel
• Higher layers are simulated before lower layers
![Page 12: High Performance Geo-Simulations](https://reader035.vdocument.in/reader035/viewer/2022081520/56816468550346895dd65257/html5/thumbnails/12.jpg)