The current challenges in the industrial aero space/Automotive CFD fields are to offer both better quality/featured products at cheaper price and produce faster pace to users. This forces designers to minimize risk, cost reduction and time to market. The cost influence of any product is large in the initial phases of the design, hence the confidence on the success of the product in the market should increase in early stage of the design cycle. This requires to conduct high fidelity flow simulations to enter the flow dynamic design process much earlier than today. Also this needs to use more advanced flow modelling levels throughout complete design process. However, the turnaround time for a simulation is a restricting factor when deciding the physical modelling level in each design phase. Parallel processing is commonly used to reduce the turnaround time and with the introduction of Linux clusters in the mid -90’s, cost efficient supercomputing resources has become available in large scale for wider community. Several 10’s of thousands of nodes Linux clusters with specific configurations to suite to each application needs have become popular. Several HPC clustering techniques and custom configurations of sub components have provided improved turnaround times which in turn providing possibility to increase the modelling capability. This session is aimed to discuss these choices and configurations and the challenges.
Confirmed speakers include:
- Vasanth Jain, Chief Scientist, GE Global Research Centre, JFWTC, Bangalore, India
- Thomas Babu/Ashok Kumar, Group Director, Vikram Sarabhai Space Centre, ISRO, Dept. Of Space, India
- Anutosh Moitra, Group head, Aero space Practise, TATA Consultancy Services, Bangalore
- Bala Krishnan, Professor, Aero space department, Indian Institute of Science, Bangalore, India