Flood Modeller has redefined the standard for speed and accuracy using the very latest in GPU acceleration techniques. Flood Modeller’s existing code will now run on NVIDIA GPUs, providing you with modelling results significantly faster than before.
Compatible with NVIDIA GPUs
Up to 94% faster when using a laptop GPU - no need to worry about using small grid sizes or expensive GPU cards
Easy to set up and use - simply select the GPU solver when you are ready to run your model
Produces highly accurate results using the same benchmarked and widely accepted 2D solvers
Multiple 2D domains can be used to represent different areas of a floodplain at different resolutions.
Run your simulations in Flood Cloud using a NVIDIA Tesla v100 GPU
Flood Modeller’s 2D GPU solver doesn’t require high-specification graphics cards and provides faster simulation runtimes using a standard laptop GPU. It allows you to improve your modelling capabilities, by constructing flood models with finer grid resolution and wider extents, enabling more accurate and detailed results.
It uses the same underlying mathematics as the original CPU solver, allowing the same outputs to be obtained, just faster. The GPU solver adopts the 2D TVD solver methodology due to its compatibility with parallelisation, which provides even faster runtimes.
The 2D GPU solver can be dynamically linked to Flood Modeller’s 1D river and 1D urban solvers enabling you to significantly reduce the time it takes to run your fully integrated models. This functionality is fully embedded within Flood Modeller's powerful GIS interface, therefore it’s extremely easy to set up and start using on your projects.
Multiple 2D domains, with different cell sizes, time-steps and simulation times can be coupled to a single 1D model to represent different areas of a floodplain at different resolutions.
When you have many models to run, you can utilise Flood Modeller's batch runner to setup a suite of simulations and manage the running of each model. You also have the option of utilising Flood Cloud.
Our latest testing of the 2D GPU solver has demonstrated runtime savings of up to 94% when using the NVIDIA Quadro T2000 GPU (typically found in a laptop) compared to a i7-9850H CPU @2.60 GHz. We have also tested it on a higher specification GPU (GeForce RTX 2080 Ti) which resulted in runtime savings of up to 98%.
Environment Agency Benchmark Test #5