SlideShare a Scribd company logo
1 of 27
1
Windsurfing with APPA: Automating Computational
Fluid Dynamics Simulations of Wind Flow using
Cloud Computing*
Anshul Jindal, Benedikt Strahm, Vladimir Podolskiy, Michael Gerndt
Chair of Computer Architecture and Parallel Systems
Technical University of Munich (TUM), Germany
Session: CISA
Parallel, Distributed, and Network-Based Processing 2020
13th March 2020
*Supported by Google Cloud Platform research
credits and AWS cloud credits.
Cloud and IoT RG @ Chair of Computer Architecture and Parallel Systems
Technical University of Munich
 Research Areas:
• Self-adaptive Cloud (predictive autoscaling for VMs).
• AI for smart Cloud operations (anomaly detection and failure predictions).
• Modelling of microservices applications and many more…
 Key Publications:
• [2019, ICPE ] A. Jindal, V. Podolskiy, M. Gerndt "Performance Modeling for Cloud Microservice Applications”.
• [2018, CLOUD] Vladimir Podolskiy, Anshul Jindal and Michael Gerndt. IaaS Reactive Autoscaling Performance
Challenges. The 2018 IEEE International Conference on Cloud Computing (CLOUD 2018).
• [2017, SC2 ] Anshul Jindal, Vladimir Podolskiy and Michael Gerndt. 2017. Multilayered Cloud Applications
Autoscaling Performance Estimation.
2
We are…
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Prof. Dr.
Michael Gerndt
Ph.D Student
Vladimir Podolskiy
Ph.D. Student
Anshul Jindal
+ Other Students
 Introduction
 Motivation
 Goals
 Related Work
 Developed Solution
 Evaluation and Results
 Conclusion and future work
3
Outline
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Introduction
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 4
 Wind loading on buildings is a potential cause for structural damages and collapses.
 Wind Engineering investigates the impact of wind flow on buildings in different
environments.
 Wind tunnel-based physical testing was established as the standard to evaluate the
impact of wind.
 Computational fluid dynamics (CFD) uses numerical analysis to conduct simulations of
turbulent wind flow around structures.
 It can serve as an alternative approach to conventional wind tunnel testing to assess the
wind flow around tall buildings.
5
Introduction
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 Computational fluid dynamics (CFD) is a branch of fluid mechanics which deals with
numerical solutions of various fluid problems.
 The flow of air around tall buildings is classified as an unsteady turbulent Newtonian
fluid nearly always in subsonic range and therefore incompressible Navier-Stokes
equation is used to model it.
 To perform numerical simulations for buildings,
a so-called virtual wind tunnel is used.
 It is a three-dimensional space that encloses the
geometry of the building and are discretized into
computational cells.
 Process of converting the flow problem from a physical statement:
6
CFD in wind engineering on tall buildings
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Physical
System
Mathematical
Model
Discrete
Model
Numerical
Approximation
Computed
Solution
Motivation
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 7
 CFD can serve as a complementary approach to conventional wind tunnel testing to
assess the wind flow around tall buildings.
 CFD simulations conventionally run on supercomputers and compute clusters using
specialized software such as OpenFOAM.
8
Motivation
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Limited availability and high maintenance
costs of supercomputers and clusters.
Use of on-demand offer of compute capacity by Cloud
Require extensive expertise and experience in working
with cloud computing.
Problem
Goals
9Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
10
Goals
Use compute capacity of the Cloud for
running simulations.
Automatic deployment and
execution of CFD simulations.
Estimate the optimal number of vCPUs
based on running time and cost.
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
1
2
3
Related Work
11Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 J. Slawinski, T. Passerini, U. Villa, A. Veneziani, and V. Sunderam,“Experiences with
target-platform heterogeneity in clouds, grids, and on-premises resources,” in2012 IEEE
26th International Parallel and Distributed Processing Symposium Workshops PhD
Forum, May 2012,pp. 41–52
• Ported two production CFD based simulations on four different platforms (one of which was
Amazon EC2)
• Claimed Cloud laaS resources can be utilized for scientific CFD simulations possibly at lower cost.
• Deployed manually due to cumbersome nature of the requirements and it took them up to a day
for the deployment only.
R. Ledyayev and H. Richter, “High performance computing in a cloud using
openstack,”Cloud Computing, pp. 108–113, 2014.
• Benchmarked OpenFOAM on OpenStack platform.
• Found that OpenStack platform might not be well-suited for running OpenFOAM based
simulations due to the performance degradation.
• Claimed certain measures like changing the hardware and running multiple tests can improve the
performance.
12
Related Work
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 O'Leary, Patrick et al. “HPCCloud: A Cloud/Web-Based Simulation Environment.” 2015
IEEE 7th International Conference on Cloud Computing Technology and Science
(CloudCom) (2015): 25-33
• Allows to create or define a simulation using the given input decks and then it submits to a cluster
in the Cloud for processing
• We have left the simulation creation to the domain expert and other stuff like deployment,
running of simulation, monitoring etc. are off-loaded from the user to the developed system.
13
Related Work Cont..
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Developed Solution
14Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
15
Overall System Architecture
Mongo
DB
Test data
saved
Resources Monitoring
data saved
Influx
DB
Output
UserInterface/APIInterface
Visualizationand
tables
Application
Configuratio
n
DeploymentConfiguration
Application
Source
Application
Deployment
AWS Cloud Interface
Google Cloud Interface
Application Deployer
Application Container
Monitoring Agent
Virtual Machine
Cloud Bucket
Test
Start
APPA
Cloud Service Provider
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
16
Automated Parallel Processing Application (APPA) Workflow
OpenFOAM simulation source code of the
building
Configuration of simulation parameters
VM “1”
Running of simulations on the Cloud
VM “2” VM “n-1” VM “n”
Simulation results storage
Analysis, visualization and download
Monitoring data storage
…
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
17
Dashboard screenshot for starting the simulation
Input Configuration ParametersChoose Cloud Service
Provider
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
18
Dashboard screenshot for viewing simulation‘s status
Type of
Instance
Download
Simulation
results
Monitor VM in real
time
Current Status of
Simulation (number of
timesteps completed)
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Evaluation
19Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 Instances equipped with 150GB of storage.
 OS used was Ubuntu 18.04-LTS-bionic
 Docker version 18.03 CE and Docker-Compose version 1.23.1 was used.
 OpenFOAM Docker image version v1812 was used.
20
Experimental Configurations (System)
Cloud
Provider
Instance Type vCPU Memory (GiB)
AWS c4.8xlarge 36 60
GCP
n1-highcpu-32 32 28.8
n1-highcpu-64 64 57.6
n1-highcpu-96 96 86.4
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 The simulated building is the Commonwealth Advisory Aeronautical Council (CAARC)
standard tall building model.
 The building geometry consists of a prismatic shape with a rectangular cross-section.
 The domain itself was decomposed into four zones with gradually refined grid sizes.
 Two different grid resolutions, both with hexahedralcells, were created:
554,542 cells for Mesh A and 1,476,020 cells for Mesh B
 Number of time steps was equal to 6,250 for each simulation.
21
Experimental Configurations (Simulation)
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
22
Performance analysis for Mesh A with different objectives
16 Cores AWS32 Cores GCP (n1-32)36 Cores GCP (n1-64) 10 Cores AWS10 Cores GCP (n1-64)12 Cores GCP (n1-32)
Objective: Reducing Average runtime per timestep Objective: Reducing Total Cost
Overall Best: 36 cores of n1-highcpu-64 (1.17s) Overall Best: 10 cores of n1-highcpu-64 (78.55 USD)
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
23
Performance analysis for Mesh B with different objectives
42 Cores GCP (n1-96) 28 Cores GCP (n1-96)
Objective: Reducing Average runtime per timestep Objective: Reducing Total Cost
Overall Best: 42 cores of n1-highcpu-96 (2.55s) Overall Best: 28 cores of n1-highcpu-96 (298.48 USD)
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
Conclusion
24Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 Cloud computing provides a valuable alternative for running CFD-simulations on-
demand.
 The introduced APPA tool provides a user-friendly interface for conducting CFD-
simulations and deployment, running of simulation, monitoring etc. tasks are off-loaded
from the user.
 Results show that:
• use of GCE instance types can result both in low total cost (78.55USD) and in low
execution time per timestep (0.77seconds).
• lowering the number of cores results in the reduction of the total cost.
25
Conclusion
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
 The monitoring of the simulation time and of the resources usage to enable the
optimization of resources by applying early stopping.
 Deriving the model that relates domain-specific parameters used for CFD-simulations to
the total cost and the execution time.
 Porting of this platform on a serverless computing framework so that individual
simulations can be run as a function.
26
Future Work
Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
27
Contact
anshul.jindal@tum.de
/ansjin
/Anshul_Jindal4
Thank you for your attentention!
Questions ?
Scan for checking the
tool on GitHub
https://github.com/Cloud-Pie/APPA

More Related Content

What's hot

Mapping & OMS Integrations with Third-Party Hardware/Software Solutions
Mapping & OMS Integrations with Third-Party Hardware/Software SolutionsMapping & OMS Integrations with Third-Party Hardware/Software Solutions
Mapping & OMS Integrations with Third-Party Hardware/Software Solutionsjasontuck
 
Increasing ROI Through Simulation and the 'Digital Twin'
Increasing ROI Through Simulation and the 'Digital Twin'Increasing ROI Through Simulation and the 'Digital Twin'
Increasing ROI Through Simulation and the 'Digital Twin'GSE Systems, Inc.
 
Computer Architecture
Computer ArchitectureComputer Architecture
Computer ArchitectureSonal Patil
 
OpenACC Monthly Highlights May 2017
OpenACC Monthly Highlights  May 2017OpenACC Monthly Highlights  May 2017
OpenACC Monthly Highlights May 2017NVIDIA
 
How to Prepare Weather and Climate Models for Future HPC Hardware
How to Prepare Weather and Climate Models for Future HPC HardwareHow to Prepare Weather and Climate Models for Future HPC Hardware
How to Prepare Weather and Climate Models for Future HPC Hardwareinside-BigData.com
 
A Comparison Study between Inferred State-Space and Neural Network Based Syst...
A Comparison Study between Inferred State-Space and Neural Network Based Syst...A Comparison Study between Inferred State-Space and Neural Network Based Syst...
A Comparison Study between Inferred State-Space and Neural Network Based Syst...Ahmed Momtaz Hosny, PhD
 
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...Seval Çapraz
 
MapR Edge : Act Locally Learn Globally
MapR Edge : Act Locally Learn GloballyMapR Edge : Act Locally Learn Globally
MapR Edge : Act Locally Learn Globallyridhav
 

What's hot (10)

Mapping & OMS Integrations with Third-Party Hardware/Software Solutions
Mapping & OMS Integrations with Third-Party Hardware/Software SolutionsMapping & OMS Integrations with Third-Party Hardware/Software Solutions
Mapping & OMS Integrations with Third-Party Hardware/Software Solutions
 
Performance_and_Cost_Evaluation
Performance_and_Cost_EvaluationPerformance_and_Cost_Evaluation
Performance_and_Cost_Evaluation
 
Increasing ROI Through Simulation and the 'Digital Twin'
Increasing ROI Through Simulation and the 'Digital Twin'Increasing ROI Through Simulation and the 'Digital Twin'
Increasing ROI Through Simulation and the 'Digital Twin'
 
LinkedIn
LinkedInLinkedIn
LinkedIn
 
Computer Architecture
Computer ArchitectureComputer Architecture
Computer Architecture
 
OpenACC Monthly Highlights May 2017
OpenACC Monthly Highlights  May 2017OpenACC Monthly Highlights  May 2017
OpenACC Monthly Highlights May 2017
 
How to Prepare Weather and Climate Models for Future HPC Hardware
How to Prepare Weather and Climate Models for Future HPC HardwareHow to Prepare Weather and Climate Models for Future HPC Hardware
How to Prepare Weather and Climate Models for Future HPC Hardware
 
A Comparison Study between Inferred State-Space and Neural Network Based Syst...
A Comparison Study between Inferred State-Space and Neural Network Based Syst...A Comparison Study between Inferred State-Space and Neural Network Based Syst...
A Comparison Study between Inferred State-Space and Neural Network Based Syst...
 
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...
GPU-Accelerated Route Planning of Multi-UAV Systems Using Simulated Annealing...
 
MapR Edge : Act Locally Learn Globally
MapR Edge : Act Locally Learn GloballyMapR Edge : Act Locally Learn Globally
MapR Edge : Act Locally Learn Globally
 

Similar to Windsurfing with APPA: Automating the Computational Fluid Dynamics Simulations of Wind Flow using the Cloud Computing

Online Memory Leak Detection in the Cloud-based Infrastructures
Online Memory Leak Detection in the Cloud-based InfrastructuresOnline Memory Leak Detection in the Cloud-based Infrastructures
Online Memory Leak Detection in the Cloud-based InfrastructuresAnshul Jindal
 
An optimized scientific workflow scheduling in cloud computing
An optimized scientific workflow scheduling in cloud computingAn optimized scientific workflow scheduling in cloud computing
An optimized scientific workflow scheduling in cloud computingDIGVIJAY SHINDE
 
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...IRJET Journal
 
CFD Apps: Presentation of the Urban Wind Study App
CFD Apps: Presentation of the Urban Wind Study AppCFD Apps: Presentation of the Urban Wind Study App
CFD Apps: Presentation of the Urban Wind Study AppJulien de Charentenay
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computingijujournal
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computingijujournal
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computingijujournal
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computingijujournal
 
Smart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudSmart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudWolfgang Gentzsch
 
NAFEMS Smart Manufacturing - UberCloud
NAFEMS Smart Manufacturing - UberCloudNAFEMS Smart Manufacturing - UberCloud
NAFEMS Smart Manufacturing - UberCloudThomas Francis
 
Smart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudSmart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudThe UberCloud
 
Cost-aware scalability of applications in public clouds
Cost-aware scalability of applications in public clouds Cost-aware scalability of applications in public clouds
Cost-aware scalability of applications in public clouds Daniel Moldovan
 
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...Francisco Marcelo Pereira Hernandez
 
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADING
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADINGCONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADING
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADINGIJCNCJournal
 

Similar to Windsurfing with APPA: Automating the Computational Fluid Dynamics Simulations of Wind Flow using the Cloud Computing (20)

Online Memory Leak Detection in the Cloud-based Infrastructures
Online Memory Leak Detection in the Cloud-based InfrastructuresOnline Memory Leak Detection in the Cloud-based Infrastructures
Online Memory Leak Detection in the Cloud-based Infrastructures
 
An optimized scientific workflow scheduling in cloud computing
An optimized scientific workflow scheduling in cloud computingAn optimized scientific workflow scheduling in cloud computing
An optimized scientific workflow scheduling in cloud computing
 
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...
Artificial Intelligence Based Method for Obtaining Aerodynamic Characteristic...
 
CFD Apps: Presentation of the Urban Wind Study App
CFD Apps: Presentation of the Urban Wind Study AppCFD Apps: Presentation of the Urban Wind Study App
CFD Apps: Presentation of the Urban Wind Study App
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computing
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computing
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computing
 
A Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud ComputingA Review on Scheduling in Cloud Computing
A Review on Scheduling in Cloud Computing
 
Fog computing
Fog computingFog computing
Fog computing
 
RMC_final
RMC_finalRMC_final
RMC_final
 
Smart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudSmart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the Cloud
 
P1151220195
P1151220195P1151220195
P1151220195
 
NAFEMS Smart Manufacturing - UberCloud
NAFEMS Smart Manufacturing - UberCloudNAFEMS Smart Manufacturing - UberCloud
NAFEMS Smart Manufacturing - UberCloud
 
Smart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the CloudSmart Manufacturing: CAE in the Cloud
Smart Manufacturing: CAE in the Cloud
 
Isncc2020
Isncc2020Isncc2020
Isncc2020
 
Cost-aware scalability of applications in public clouds
Cost-aware scalability of applications in public clouds Cost-aware scalability of applications in public clouds
Cost-aware scalability of applications in public clouds
 
Cloud Computing
Cloud ComputingCloud Computing
Cloud Computing
 
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...
Monitoring and Forecasting of Air Emissions with IoT Measuring Stations and a...
 
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADING
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADINGCONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADING
CONTEXT-AWARE DECISION MAKING SYSTEM FOR MOBILE CLOUD OFFLOADING
 
51 b wittmer_latest_features_of_p_vsyst
51 b wittmer_latest_features_of_p_vsyst51 b wittmer_latest_features_of_p_vsyst
51 b wittmer_latest_features_of_p_vsyst
 

Recently uploaded

(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service
(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service
(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Serviceranjana rawat
 
100-Concepts-of-AI by Anupama Kate .pptx
100-Concepts-of-AI by Anupama Kate .pptx100-Concepts-of-AI by Anupama Kate .pptx
100-Concepts-of-AI by Anupama Kate .pptxAnupama Kate
 
Smarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxSmarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxolyaivanovalion
 
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptx
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptxBPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptx
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptxMohammedJunaid861692
 
Mature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxMature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxolyaivanovalion
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfMarinCaroMartnezBerg
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysismanisha194592
 
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...amitlee9823
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxfirstjob4
 
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...shivangimorya083
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfRachmat Ramadhan H
 
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfAccredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfadriantubila
 
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Delhi Call girls
 
Data-Analysis for Chicago Crime Data 2023
Data-Analysis for Chicago Crime Data  2023Data-Analysis for Chicago Crime Data  2023
Data-Analysis for Chicago Crime Data 2023ymrp368
 
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAl Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAroojKhan71
 
Generative AI on Enterprise Cloud with NiFi and Milvus
Generative AI on Enterprise Cloud with NiFi and MilvusGenerative AI on Enterprise Cloud with NiFi and Milvus
Generative AI on Enterprise Cloud with NiFi and MilvusTimothy Spann
 
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip CallDelhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Callshivangimorya083
 

Recently uploaded (20)

(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service
(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service
(PARI) Call Girls Wanowrie ( 7001035870 ) HI-Fi Pune Escorts Service
 
100-Concepts-of-AI by Anupama Kate .pptx
100-Concepts-of-AI by Anupama Kate .pptx100-Concepts-of-AI by Anupama Kate .pptx
100-Concepts-of-AI by Anupama Kate .pptx
 
Smarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxSmarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptx
 
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptx
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptxBPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptx
BPAC WITH UFSBI GENERAL PRESENTATION 18_05_2017-1.pptx
 
Mature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxMature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptx
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdf
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
 
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore...
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptx
 
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
 
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfAccredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
 
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get CytotecAbortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
 
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
 
Data-Analysis for Chicago Crime Data 2023
Data-Analysis for Chicago Crime Data  2023Data-Analysis for Chicago Crime Data  2023
Data-Analysis for Chicago Crime Data 2023
 
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in Kishangarh
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in  KishangarhDelhi 99530 vip 56974 Genuine Escort Service Call Girls in  Kishangarh
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in Kishangarh
 
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
 
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAl Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
 
Generative AI on Enterprise Cloud with NiFi and Milvus
Generative AI on Enterprise Cloud with NiFi and MilvusGenerative AI on Enterprise Cloud with NiFi and Milvus
Generative AI on Enterprise Cloud with NiFi and Milvus
 
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip CallDelhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls CP 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
 

Windsurfing with APPA: Automating the Computational Fluid Dynamics Simulations of Wind Flow using the Cloud Computing

  • 1. 1 Windsurfing with APPA: Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing* Anshul Jindal, Benedikt Strahm, Vladimir Podolskiy, Michael Gerndt Chair of Computer Architecture and Parallel Systems Technical University of Munich (TUM), Germany Session: CISA Parallel, Distributed, and Network-Based Processing 2020 13th March 2020 *Supported by Google Cloud Platform research credits and AWS cloud credits.
  • 2. Cloud and IoT RG @ Chair of Computer Architecture and Parallel Systems Technical University of Munich  Research Areas: • Self-adaptive Cloud (predictive autoscaling for VMs). • AI for smart Cloud operations (anomaly detection and failure predictions). • Modelling of microservices applications and many more…  Key Publications: • [2019, ICPE ] A. Jindal, V. Podolskiy, M. Gerndt "Performance Modeling for Cloud Microservice Applications”. • [2018, CLOUD] Vladimir Podolskiy, Anshul Jindal and Michael Gerndt. IaaS Reactive Autoscaling Performance Challenges. The 2018 IEEE International Conference on Cloud Computing (CLOUD 2018). • [2017, SC2 ] Anshul Jindal, Vladimir Podolskiy and Michael Gerndt. 2017. Multilayered Cloud Applications Autoscaling Performance Estimation. 2 We are… Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 Prof. Dr. Michael Gerndt Ph.D Student Vladimir Podolskiy Ph.D. Student Anshul Jindal + Other Students
  • 3.  Introduction  Motivation  Goals  Related Work  Developed Solution  Evaluation and Results  Conclusion and future work 3 Outline Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 4. Introduction Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 4
  • 5.  Wind loading on buildings is a potential cause for structural damages and collapses.  Wind Engineering investigates the impact of wind flow on buildings in different environments.  Wind tunnel-based physical testing was established as the standard to evaluate the impact of wind.  Computational fluid dynamics (CFD) uses numerical analysis to conduct simulations of turbulent wind flow around structures.  It can serve as an alternative approach to conventional wind tunnel testing to assess the wind flow around tall buildings. 5 Introduction Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 6.  Computational fluid dynamics (CFD) is a branch of fluid mechanics which deals with numerical solutions of various fluid problems.  The flow of air around tall buildings is classified as an unsteady turbulent Newtonian fluid nearly always in subsonic range and therefore incompressible Navier-Stokes equation is used to model it.  To perform numerical simulations for buildings, a so-called virtual wind tunnel is used.  It is a three-dimensional space that encloses the geometry of the building and are discretized into computational cells.  Process of converting the flow problem from a physical statement: 6 CFD in wind engineering on tall buildings Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 Physical System Mathematical Model Discrete Model Numerical Approximation Computed Solution
  • 7. Motivation Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 7
  • 8.  CFD can serve as a complementary approach to conventional wind tunnel testing to assess the wind flow around tall buildings.  CFD simulations conventionally run on supercomputers and compute clusters using specialized software such as OpenFOAM. 8 Motivation Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 Limited availability and high maintenance costs of supercomputers and clusters. Use of on-demand offer of compute capacity by Cloud Require extensive expertise and experience in working with cloud computing. Problem
  • 9. Goals 9Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 10. 10 Goals Use compute capacity of the Cloud for running simulations. Automatic deployment and execution of CFD simulations. Estimate the optimal number of vCPUs based on running time and cost. Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020 1 2 3
  • 11. Related Work 11Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 12.  J. Slawinski, T. Passerini, U. Villa, A. Veneziani, and V. Sunderam,“Experiences with target-platform heterogeneity in clouds, grids, and on-premises resources,” in2012 IEEE 26th International Parallel and Distributed Processing Symposium Workshops PhD Forum, May 2012,pp. 41–52 • Ported two production CFD based simulations on four different platforms (one of which was Amazon EC2) • Claimed Cloud laaS resources can be utilized for scientific CFD simulations possibly at lower cost. • Deployed manually due to cumbersome nature of the requirements and it took them up to a day for the deployment only. R. Ledyayev and H. Richter, “High performance computing in a cloud using openstack,”Cloud Computing, pp. 108–113, 2014. • Benchmarked OpenFOAM on OpenStack platform. • Found that OpenStack platform might not be well-suited for running OpenFOAM based simulations due to the performance degradation. • Claimed certain measures like changing the hardware and running multiple tests can improve the performance. 12 Related Work Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 13.  O'Leary, Patrick et al. “HPCCloud: A Cloud/Web-Based Simulation Environment.” 2015 IEEE 7th International Conference on Cloud Computing Technology and Science (CloudCom) (2015): 25-33 • Allows to create or define a simulation using the given input decks and then it submits to a cluster in the Cloud for processing • We have left the simulation creation to the domain expert and other stuff like deployment, running of simulation, monitoring etc. are off-loaded from the user to the developed system. 13 Related Work Cont.. Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 14. Developed Solution 14Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 15. 15 Overall System Architecture Mongo DB Test data saved Resources Monitoring data saved Influx DB Output UserInterface/APIInterface Visualizationand tables Application Configuratio n DeploymentConfiguration Application Source Application Deployment AWS Cloud Interface Google Cloud Interface Application Deployer Application Container Monitoring Agent Virtual Machine Cloud Bucket Test Start APPA Cloud Service Provider Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 16. 16 Automated Parallel Processing Application (APPA) Workflow OpenFOAM simulation source code of the building Configuration of simulation parameters VM “1” Running of simulations on the Cloud VM “2” VM “n-1” VM “n” Simulation results storage Analysis, visualization and download Monitoring data storage … Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 17. 17 Dashboard screenshot for starting the simulation Input Configuration ParametersChoose Cloud Service Provider Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 18. 18 Dashboard screenshot for viewing simulation‘s status Type of Instance Download Simulation results Monitor VM in real time Current Status of Simulation (number of timesteps completed) Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 19. Evaluation 19Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 20.  Instances equipped with 150GB of storage.  OS used was Ubuntu 18.04-LTS-bionic  Docker version 18.03 CE and Docker-Compose version 1.23.1 was used.  OpenFOAM Docker image version v1812 was used. 20 Experimental Configurations (System) Cloud Provider Instance Type vCPU Memory (GiB) AWS c4.8xlarge 36 60 GCP n1-highcpu-32 32 28.8 n1-highcpu-64 64 57.6 n1-highcpu-96 96 86.4 Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 21.  The simulated building is the Commonwealth Advisory Aeronautical Council (CAARC) standard tall building model.  The building geometry consists of a prismatic shape with a rectangular cross-section.  The domain itself was decomposed into four zones with gradually refined grid sizes.  Two different grid resolutions, both with hexahedralcells, were created: 554,542 cells for Mesh A and 1,476,020 cells for Mesh B  Number of time steps was equal to 6,250 for each simulation. 21 Experimental Configurations (Simulation) Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 22. 22 Performance analysis for Mesh A with different objectives 16 Cores AWS32 Cores GCP (n1-32)36 Cores GCP (n1-64) 10 Cores AWS10 Cores GCP (n1-64)12 Cores GCP (n1-32) Objective: Reducing Average runtime per timestep Objective: Reducing Total Cost Overall Best: 36 cores of n1-highcpu-64 (1.17s) Overall Best: 10 cores of n1-highcpu-64 (78.55 USD) Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 23. 23 Performance analysis for Mesh B with different objectives 42 Cores GCP (n1-96) 28 Cores GCP (n1-96) Objective: Reducing Average runtime per timestep Objective: Reducing Total Cost Overall Best: 42 cores of n1-highcpu-96 (2.55s) Overall Best: 28 cores of n1-highcpu-96 (298.48 USD) Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 24. Conclusion 24Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 25.  Cloud computing provides a valuable alternative for running CFD-simulations on- demand.  The introduced APPA tool provides a user-friendly interface for conducting CFD- simulations and deployment, running of simulation, monitoring etc. tasks are off-loaded from the user.  Results show that: • use of GCE instance types can result both in low total cost (78.55USD) and in low execution time per timestep (0.77seconds). • lowering the number of cores results in the reduction of the total cost. 25 Conclusion Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 26.  The monitoring of the simulation time and of the resources usage to enable the optimization of resources by applying early stopping.  Deriving the model that relates domain-specific parameters used for CFD-simulations to the total cost and the execution time.  Porting of this platform on a serverless computing framework so that individual simulations can be run as a function. 26 Future Work Anshul Jindal |Windsurfing with APPA : Automating Computational Fluid Dynamics Simulations of Wind Flow using Cloud Computing | PDP 2020 | 13th March 2020
  • 27. 27 Contact anshul.jindal@tum.de /ansjin /Anshul_Jindal4 Thank you for your attentention! Questions ? Scan for checking the tool on GitHub https://github.com/Cloud-Pie/APPA