SlideShare uma empresa Scribd logo
1 de 15
Applied Math Modeling Inc Data Center Airflow Modeling Tools 1
Applied Math Modeling Inc Founded in 2008 as a spin-out of the ANSYS Corporation. Responsible for Development, Sales, & Support for the CoolSim Data Center Modeling application through partnership with ANSYS Inc. Focused on delivering cost-effective engineering simulation tools using a SaaS model of delivery & support. Staffed by industry experts in the field of CFD sales, development, and support. 2
Data Center Market Overview EPA report issued in August of 2007: 1.5% of total US electricity is consumed by data centers 61BkWh Cost of electricity $4.5b Expected to grow to 100BkWh within 5 years ($7.4B) Among the key recommendations, improved airflow management is the most cost effective measure yielding up to 30% energy reduction in operational costs. “Optimize data center airflow configuration using CFD modeling and visualization tools” 3
Data Center Energy Consumption Dominant factors are IT Power and Cooling Power. IT Load, CRAC/Chiller make up 75% of total  UPS, PDU, Switchgear are all 2nd order effects Source: Uptime Institute 4
Benefits of Data Center Airflow Modeling	 The user creates a model of the data center using CoolSim for performing “what if” scenarios of potential changes What is the effect of adding servers to the data center? What happens if a cooling unit(s) fails? What is the effect of hot aisle/cold aisle containment? What is the maximum IT load a given data center can handle? Allows users to predict the and temperature anywhere within the data center and make informed decisions on cooling or IT equipment placement. Provides the necessary information for significant cost savings in new data center designs and/or reduce operational costs of existing data centers. 5
How Does CoolSim Work? CoolSim incorporates ANSYS Fluent CFD technology Same software used in Aerospace and Automotive All meshing, solving, and post processing are performed on a remote HPC cluster. CoolSim uses a subscription based SaaS (Software as a Service) model where price is correlated to anticipated annual use. 6
How Does CoolSim Work? CoolSim is a highly automated tool User builds a model of a data center using easy to use GUI Includes libraries of common cooling equipment and equipment racks, and IT equipment Model is automatically meshed, solved, and post processed  Embedded knowledge drawing upon 20 years of CFD modeling experience from aerospace and automotive markets Results are presented in 3D HTML output reports Eliminates costly “human time” in post processing Eliminates “human error” in post processing output plots User learning curve measured in hours Run times average less than an hour Multiple jobs can be run concurrently 7
CoolSim Key Features Modeling Capability Fully coupled CFD analysis powered by ANSYS CFD Solvers Model raised and non-raised floor data centers Models in-row and overhead supplemental cooling solutions Server specific rack modeling with gaps and blanking Ability to model supply ducts and diffusers Ability to model under-floor and overhead cable trays, under-floor pipes, baffles, blocks, PDU’s, etc Ability to model ducted ceiling return data centers Ability to model CRAC failure scenarios efficiently and quickly. 8
CoolSim Key Features Output Capability Thermal maps at rack inlets, outlets and throughout data center. Detailed airflow patterns in the room showing return airflow path of hot air revealing any mixing or short circuiting currents. CRAC performance report with return and supply air temperature and relative load balancing on various units. Rack by rack cooling performance evaluation and compliance with the ASHRAE recommended thermal guidelines. Airflow patterns and static pressure distribution in supply plenum. Airflow rate (CFM) prediction for each vent tile and cable cutout 3D models of all visual output with high quality 2D image output. Animations of supply and return air path lines, temperature planes in X, Y and Z, and isothermal surfaces @ 2 degree intervals. 9
3D Rack Thermal Maps
Performing “What-if” Studies Can we reduce inlet temperatures by simply adding low cost baffles to reduce rack recirculation?
Results of Adding Baffles to Racks By adding simple baffles to racks, recirculation is reduced and server inlet temperatures drop.
CoolSim Automated Output Reports Allows the easy comparison of multiple scenarios 13
Summary The density of computer equipment will continue to increase compounding cooling problems. Typical data centers draw 2X more power than required by the IT equipment alone for cooling the computer equipment. Increasing the thermal efficiency of the data center can create savings from 30-70% per year in operational costs. Using CoolSim CFD based modeling as a means to predict potential improvements Data Center efficiency is very cost effective and necessary as part of the design or modification process. The potential electricity cost savings from a 1MW data center are in the range of $2-4M over a 10 year period. 14
For More Information or a Product Demonstration  For a quote or pricing information please contact: ,[object Object],Jennifer.Beliveau@CoolSimSoftware.com 603-632-3026 For more product related information please visit: ,[object Object],15

Mais conteúdo relacionado

Semelhante a CoolSim Overview

Thinking Outside the Box – Server Design for #DataCenter Optimization
Thinking Outside the Box – Server Design for #DataCenter OptimizationThinking Outside the Box – Server Design for #DataCenter Optimization
Thinking Outside the Box – Server Design for #DataCenter OptimizationRobert Schmidt
 
Optimize power and cooling final 1
Optimize power and cooling final 1Optimize power and cooling final 1
Optimize power and cooling final 1Viridity Software
 
Optimize power and cooling final
Optimize power and cooling finalOptimize power and cooling final
Optimize power and cooling finalViridity Software
 
Applying Cloud Techniques to Address Complexity in HPC System Integrations
Applying Cloud Techniques to Address Complexity in HPC System IntegrationsApplying Cloud Techniques to Address Complexity in HPC System Integrations
Applying Cloud Techniques to Address Complexity in HPC System Integrationsinside-BigData.com
 
Green IT @ STRATO - Rene Weinholtz
Green IT @ STRATO - Rene WeinholtzGreen IT @ STRATO - Rene Weinholtz
Green IT @ STRATO - Rene Weinholtzcatherinewall
 
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...IBMAsean
 
Wp a-break-in-the-clouds
Wp a-break-in-the-cloudsWp a-break-in-the-clouds
Wp a-break-in-the-cloudsMohsen Tayefeh
 
IRJET- Cloud Cost Analyzer and Optimizer
IRJET- Cloud Cost Analyzer and OptimizerIRJET- Cloud Cost Analyzer and Optimizer
IRJET- Cloud Cost Analyzer and OptimizerIRJET Journal
 
Data Engine for NoSQL - IBM Power Systems
Data Engine for NoSQL - IBM Power SystemsData Engine for NoSQL - IBM Power Systems
Data Engine for NoSQL - IBM Power SystemsthinkASG
 
Optimizing The Data Centre Environment
Optimizing The Data Centre EnvironmentOptimizing The Data Centre Environment
Optimizing The Data Centre Environmentmixalisg
 
Ibm low energy_final low res version
Ibm low energy_final low res versionIbm low energy_final low res version
Ibm low energy_final low res versionEstelle Andlauer
 
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...Principled Technologies
 
The need for a fluid data centre, what that actually means, and the hurdles i...
The need for a fluid data centre, what that actually means, and the hurdles i...The need for a fluid data centre, what that actually means, and the hurdles i...
The need for a fluid data centre, what that actually means, and the hurdles i...Nordic Digital Business Summit
 
How Accurate is Future Facilities 6Sigma DCX
How Accurate is Future Facilities 6Sigma DCXHow Accurate is Future Facilities 6Sigma DCX
How Accurate is Future Facilities 6Sigma DCXRobert Schmidt
 
Schneider - ASK THE EXPERT - ITNEXT
Schneider - ASK THE EXPERT - ITNEXTSchneider - ASK THE EXPERT - ITNEXT
Schneider - ASK THE EXPERT - ITNEXTAniket Patange
 
Green cloud computing
Green cloud computingGreen cloud computing
Green cloud computingNalini Mehta
 

Semelhante a CoolSim Overview (20)

Thinking Outside the Box – Server Design for #DataCenter Optimization
Thinking Outside the Box – Server Design for #DataCenter OptimizationThinking Outside the Box – Server Design for #DataCenter Optimization
Thinking Outside the Box – Server Design for #DataCenter Optimization
 
Optimize power and cooling final 1
Optimize power and cooling final 1Optimize power and cooling final 1
Optimize power and cooling final 1
 
Optimize power and cooling final
Optimize power and cooling finalOptimize power and cooling final
Optimize power and cooling final
 
Applying Cloud Techniques to Address Complexity in HPC System Integrations
Applying Cloud Techniques to Address Complexity in HPC System IntegrationsApplying Cloud Techniques to Address Complexity in HPC System Integrations
Applying Cloud Techniques to Address Complexity in HPC System Integrations
 
Green IT @ STRATO - Rene Weinholtz
Green IT @ STRATO - Rene WeinholtzGreen IT @ STRATO - Rene Weinholtz
Green IT @ STRATO - Rene Weinholtz
 
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...
Green & Beyond: Data Center Actions to Increase Business Responsiveness and R...
 
Green It
Green ItGreen It
Green It
 
Wp a-break-in-the-clouds
Wp a-break-in-the-cloudsWp a-break-in-the-clouds
Wp a-break-in-the-clouds
 
IRJET- Cloud Cost Analyzer and Optimizer
IRJET- Cloud Cost Analyzer and OptimizerIRJET- Cloud Cost Analyzer and Optimizer
IRJET- Cloud Cost Analyzer and Optimizer
 
Data Engine for NoSQL - IBM Power Systems
Data Engine for NoSQL - IBM Power SystemsData Engine for NoSQL - IBM Power Systems
Data Engine for NoSQL - IBM Power Systems
 
Optimizing The Data Centre Environment
Optimizing The Data Centre EnvironmentOptimizing The Data Centre Environment
Optimizing The Data Centre Environment
 
Ibm low energy_final low res version
Ibm low energy_final low res versionIbm low energy_final low res version
Ibm low energy_final low res version
 
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...
Compute intensive performance efficiency comparison: HP Moonshot with AMD APU...
 
Build Energy Saving into Your Datacenter
Build Energy Saving into Your DatacenterBuild Energy Saving into Your Datacenter
Build Energy Saving into Your Datacenter
 
The need for a fluid data centre, what that actually means, and the hurdles i...
The need for a fluid data centre, what that actually means, and the hurdles i...The need for a fluid data centre, what that actually means, and the hurdles i...
The need for a fluid data centre, what that actually means, and the hurdles i...
 
How Accurate is Future Facilities 6Sigma DCX
How Accurate is Future Facilities 6Sigma DCXHow Accurate is Future Facilities 6Sigma DCX
How Accurate is Future Facilities 6Sigma DCX
 
Data center m&e
Data center  m&eData center  m&e
Data center m&e
 
Schneider - ASK THE EXPERT - ITNEXT
Schneider - ASK THE EXPERT - ITNEXTSchneider - ASK THE EXPERT - ITNEXT
Schneider - ASK THE EXPERT - ITNEXT
 
Green cloud computing
Green cloud computingGreen cloud computing
Green cloud computing
 
Green IT
Green IT Green IT
Green IT
 

CoolSim Overview

  • 1. Applied Math Modeling Inc Data Center Airflow Modeling Tools 1
  • 2. Applied Math Modeling Inc Founded in 2008 as a spin-out of the ANSYS Corporation. Responsible for Development, Sales, & Support for the CoolSim Data Center Modeling application through partnership with ANSYS Inc. Focused on delivering cost-effective engineering simulation tools using a SaaS model of delivery & support. Staffed by industry experts in the field of CFD sales, development, and support. 2
  • 3. Data Center Market Overview EPA report issued in August of 2007: 1.5% of total US electricity is consumed by data centers 61BkWh Cost of electricity $4.5b Expected to grow to 100BkWh within 5 years ($7.4B) Among the key recommendations, improved airflow management is the most cost effective measure yielding up to 30% energy reduction in operational costs. “Optimize data center airflow configuration using CFD modeling and visualization tools” 3
  • 4. Data Center Energy Consumption Dominant factors are IT Power and Cooling Power. IT Load, CRAC/Chiller make up 75% of total UPS, PDU, Switchgear are all 2nd order effects Source: Uptime Institute 4
  • 5. Benefits of Data Center Airflow Modeling The user creates a model of the data center using CoolSim for performing “what if” scenarios of potential changes What is the effect of adding servers to the data center? What happens if a cooling unit(s) fails? What is the effect of hot aisle/cold aisle containment? What is the maximum IT load a given data center can handle? Allows users to predict the and temperature anywhere within the data center and make informed decisions on cooling or IT equipment placement. Provides the necessary information for significant cost savings in new data center designs and/or reduce operational costs of existing data centers. 5
  • 6. How Does CoolSim Work? CoolSim incorporates ANSYS Fluent CFD technology Same software used in Aerospace and Automotive All meshing, solving, and post processing are performed on a remote HPC cluster. CoolSim uses a subscription based SaaS (Software as a Service) model where price is correlated to anticipated annual use. 6
  • 7. How Does CoolSim Work? CoolSim is a highly automated tool User builds a model of a data center using easy to use GUI Includes libraries of common cooling equipment and equipment racks, and IT equipment Model is automatically meshed, solved, and post processed Embedded knowledge drawing upon 20 years of CFD modeling experience from aerospace and automotive markets Results are presented in 3D HTML output reports Eliminates costly “human time” in post processing Eliminates “human error” in post processing output plots User learning curve measured in hours Run times average less than an hour Multiple jobs can be run concurrently 7
  • 8. CoolSim Key Features Modeling Capability Fully coupled CFD analysis powered by ANSYS CFD Solvers Model raised and non-raised floor data centers Models in-row and overhead supplemental cooling solutions Server specific rack modeling with gaps and blanking Ability to model supply ducts and diffusers Ability to model under-floor and overhead cable trays, under-floor pipes, baffles, blocks, PDU’s, etc Ability to model ducted ceiling return data centers Ability to model CRAC failure scenarios efficiently and quickly. 8
  • 9. CoolSim Key Features Output Capability Thermal maps at rack inlets, outlets and throughout data center. Detailed airflow patterns in the room showing return airflow path of hot air revealing any mixing or short circuiting currents. CRAC performance report with return and supply air temperature and relative load balancing on various units. Rack by rack cooling performance evaluation and compliance with the ASHRAE recommended thermal guidelines. Airflow patterns and static pressure distribution in supply plenum. Airflow rate (CFM) prediction for each vent tile and cable cutout 3D models of all visual output with high quality 2D image output. Animations of supply and return air path lines, temperature planes in X, Y and Z, and isothermal surfaces @ 2 degree intervals. 9
  • 11. Performing “What-if” Studies Can we reduce inlet temperatures by simply adding low cost baffles to reduce rack recirculation?
  • 12. Results of Adding Baffles to Racks By adding simple baffles to racks, recirculation is reduced and server inlet temperatures drop.
  • 13. CoolSim Automated Output Reports Allows the easy comparison of multiple scenarios 13
  • 14. Summary The density of computer equipment will continue to increase compounding cooling problems. Typical data centers draw 2X more power than required by the IT equipment alone for cooling the computer equipment. Increasing the thermal efficiency of the data center can create savings from 30-70% per year in operational costs. Using CoolSim CFD based modeling as a means to predict potential improvements Data Center efficiency is very cost effective and necessary as part of the design or modification process. The potential electricity cost savings from a 1MW data center are in the range of $2-4M over a 10 year period. 14
  • 15.