Super-scale CFD on GPU at SuperComputing conference in Dallas
The Zenotech team are raring to head to the Supercomputing conference in Dallas (11-16 November 2018) to share our findings on industrial scale CFD on GPU on Amazon’s AWS Stand. Our co-director, Jamil Appa, will be presenting on the performance advantages that we’ve identified when using GPUs and the resultant impact on customers.
zCFD, our computational fluid dynamics solver, has been engineered to take full advantage of the latest developments in GPU technology, and this session will showcase the industrial problems that have been solved in a fast and cost-effective manner using a combination of zCFD and the P100 and V100 GPUs available on Amazon’s AWS. To keep the session relevant, Jamil will draw on real-world examples from our customers in the aerospace, renewable energy and automotive sectors.
Traditionally these cases have been run on in-house parallel computing clusters but the larger number of GPUs per node with AWS has enabled the solving of large CFD problems in just a single instance. Benchmarking with zCFD demonstrates that a single P3 node is providing the equivalent performance to over 1100 CPU cores. As well as performance benefits, the spot market and on-demand nature of AWS can provide cost savings for customers and therefore open up a scale of simulation previously not affordable.
The session will also demonstrate how our EPIC platform makes the combination of zCFD and AWS GPUs a simple and cost-effective solution for engineers.
Find out more about the International Conference for HPC, Networking, Storage, and Analysis.
If you’re heading to SC18 this November, we hope you can make our session and drop by the AWS Stand for a chat on 3447.