hpc and it issues - ansys · deployment challenges and trends cloud computing cloud = remote access...
TRANSCRIPT
© 2011 ANSYS, Inc. September 1, 20111
HPC and IT Issues
1:00 – 4:00 PM
Wednesday, August 31
ANSYS User Conference
Houston, TX
© 2011 ANSYS, Inc. September 1, 20112
Deployment of Simulation (Trends and Issues Impacting IT)
– Discussion
Mapping HPC to Performance (Scaling, Technology Advances)
– Discussion
Optimizing IT for Remote Access (Private / Public Cloud Computing)
– Discussion
Break
Let’s Talk about Hardware (Specifying HW for ANSYS)
– Discussion
ANSYS Roadmap (Toward aligned IT Planning)
Tomorrow
1:1 Meeting Time
HPC and IT Issues
Session Agenda
© 2011 ANSYS, Inc. September 1, 20113
Deployment of Simulation (Trends and Issues Impacting IT)
– Discussion
Mapping HPC to Performance (Scaling, Technology Advances)
– Discussion
Optimizing IT for Remote Access (Private / Public Cloud Computing)
– Discussion
Break
Let’s Talk about Hardware (Specifying HW for ANSYS)
– Discussion
ANSYS Roadmap (Toward aligned IT Planning)
Tomorrow
1:1 Meeting Time
HPC and IT Issues
Session Agenda
© 2011 ANSYS, Inc. September 1, 20114
Deployment of Simulation Trends and Challenges Related to IT
© 2011 ANSYS, Inc. September 1, 20115
Outline
ANSYS and High Performance Computing • Value to Engineering
Deployment Challenges and Trends• Extreme scale up and scale out
• Enterprise deployment / remote access
• Data management
• Connection to PLM
• Cloud computing
© 2011 ANSYS, Inc. September 1, 20116
Why ANSYS Users Need HPCInsight you can’t get any other way
HPC enables high-fidelity • Include details - for reliable results
• Be sure your design is “right”
• Innovate with confidence
HPC delivers throughput• Consider multiple design ideas
• Optimize the design
• Ensure performance across range of conditions
© 2011 ANSYS, Inc. September 1, 20117
Deployment Challenges and Trends
Infrastructure for Extreme Fidelity
Simulation allows engineering to know, not guess – but only if IT can deliver dramatically scaled up infrastructure for “mega” simulations.
IT asked to deliver • Memory capacity for 1B cell models (1 TB RAM)
• 1000s of cores per mega simulation
• Data storage for files in excess of 1 Tbyte
• Scale-out licensing (e.g. ANSYS HPC Packs)
LES Fluid Mixing 160M cell model
1000 core HPCLinköping University, SWEDEN
© 2011 ANSYS, Inc. September 1, 20118
Deployment Challenges and Trends
Infrastructure for Design Exploration
Product integrity requirements imply using simulation for robust design and optimization, which increases simulation workload by 100x
IT implication is profound “scale out”• Access to 100x HW capacity, perhaps intermittently
• Job automation
• Failure recovery and high availability
• Network bandwidth and extreme data load
• Licensing for 100x workloads
© 2011 ANSYS, Inc. September 1, 20119
Deployment Challenges and Trends
Enterprise Deployment of Simulation
Local computing infrastructure for simulation is being replaced by centralized HPC resources, shared by a globally distributed workforce.
HPC
Users
© 2011 ANSYS, Inc. September 1, 201110
Deployment Challenges and Trends
Enterprise Deployment of Simulation
Local computing infrastructure for simulation is being replaced by centralized HPC resources, shared by a globally distributed workforce.
IT-related challenges• Fail-proof enterprise licensing
• Controlled SW version updates; backward compatibility
• Job scheduling/prioritization
• Remote interactive access and remote visualization tools
● Critical, given file size and transfer time
• Data sharing by distributed work teams; data management
© 2011 ANSYS, Inc. September 1, 201111
Deployment Challenges and Trends
Simulation Data Management and Storage
Access to simulation results is required over the full product lifecycle. Re-use of previous simulations is important for engineering efficiency. Regulation and liability imply long-term archival storage.
IT needs to provide• Long-term storage and backup
• System for cataloging and finding simulation files for re-use
• Ability to relate simulation files and master models in PLM system
© 2011 ANSYS, Inc. September 1, 201112
Simulation Data Management
ANSYS EKM
Web Browser (IE,
Firefox, etc.)
Firewall
Desktop Application (ANSYS
Workbench, VB, etc.) File Server(http, ftp)
Relational Database(Mysql, Oracle, DB2, etc.)
Compute Cluster
Application Server(Jboss , etc.)
Content Management Repository (Jackrabbit, etc.)
ANSYS EKM
Stores
metadata
Executes simulations
and generates data
Repository of all archived files and applications
© 2011 ANSYS, Inc. September 1, 201113
Deployment Challenges and Trends
Connecting Simulation to PLM / PDM
Pull functional requirements or geometry from PLM • Maintain version control between simulation and PLM (e.g., EKM updates)
• Simulation may precede geometry
Return simulation results to PLM• Associated to the master model , with full data storage likely outside PLM
ANSYS EKM Windchill
Teamcenter
Matrix
One
Sm
arte
am
EKM Datalink
Functional SpecGeometry, Attributes
Simulation Report
© 2011 ANSYS, Inc. September 1, 201114
Deployment Challenges and Trends
Cloud Computing
Cloud = remote access to elastic infrastructure• Scale up compute capacity and data storage, intermittent workloads
• Private cloud (internal elastic provisioning) or outsourced HPC on public cloud
But challenges remain for• Remote workflow (file transfer, visualization, remote interactive access)
• IP protection, export compliance and data sharing
• Optimized use of on-premise vs. “remote” systems
ANSYS focus• Improved remote simulation workflow
• Enable / encourage seamless use of remote (or 3P hosted) hardware
● “Bring Your Own License” model
© 2011 ANSYS, Inc. September 1, 201115
“Take Home” Points / Discussion
IT infrastructure is “mission critical” for extracting high value from simulation
● Key emerging strategy topic for ANSYS and for our customers
Scalability for next generation simulation workloads is key
● Scale Up and/or Scale Out
Enterprise deployment and cloud computing share technical challenges
● Remote interactive access and visualization, file transfer, IP protection
Data management strategy is critical for effective collaboration and re-use
● Connection to PLM, where applicable
Comments / Questions / Discussion