HPC Introduction: Difference between revisions
Content deleted Content added
Stephen.kime (talk | contribs) No edit summary |
Stephen.kime (talk | contribs) No edit summary |
||
Line 9:
HPC cell discretisation is the same as TUFLOW Classic This has removed the “checkerboarding effect” observed at times in TUFLOW GPU result outputs. This means that the water levels (i.e. cell mass) is calculated at the cell centre, while velocity (i.e. momentum) is calculated at the cell mid-sides as shown in the figures below.<br>
In general most of the functionality and features of TUFLOW Classic are available in HPC, with more of these features becoming available to HPC in due time.<br>
So what does this mean for modellers? By providing the ability to run models on Graphics Cards or across multiple CPU cores, we can achieve significantly greater model run times, increasing our modelling capabilities to be able to run continuous hydraulic models, with increased cell resolution, across larger extents and more scenarios (Monte Carlo, rainfall ensembles). If you would like to learn more about HPC’s run speeds and modelling benefits then please refer to our <font color="blue"><tt> Software Benchmarking </tt></font> wiki page.<br>
*Cuda Cores are cores written to a specific language (CUDA) which is patented to Nvidia graphics cards. This means that TUFLOW is limited to running on CUDA enabled graphics cards only.<br>
| |||