HPC FAQ: Difference between revisions
Content deleted Content added
Stephen.kime (talk | contribs) Created page with " '''Q. Does the HPC 2D solver produce the same results as the Classic 2D solver? '''<br> No, TUFLOW Classic uses a 2nd order ADI (Alternating Direction Implicit) finite diffe..." |
Stephen.kime (talk | contribs) No edit summary |
||
Line 25:
If running on the same CPU hardware, a well-constructed Classic model on a good timestep is nearly always faster than HPC running on a single CPU thread. Running a single HPC simulation across multiple CPU threads may produce a faster simulation than Classic. Primary reasons why the HPC may produce a slower run, especially are discussed below. <br>
'''Q.
Trying to run multiple HPC simulations across the same CPU threads. If, for example, you have 4 CPU threads on your computer and you run two simulations that both request 4 threads, then effectively you are overloading the CPU hardware by requesting 8 threads in total. This will slow down the simulations by an approximate factor of 2.
Line 35:
Note: To request the maximum number of threads on a machine use <font color="blue"><tt>CPU Threads == MAX</tt></font> if hyperthreading is deactivated and <font color="blue"><tt>CPU Threads == MAX/2</tt></font> if hyperthreading is active. The latter requests half the number of threads, which for most Windows machines is the same as the number of physical cores. <br>
'''Q.
If running a simulation using a low end or old GPU device (these are usually the graphics card that comes standard with computer), simulations can be only marginally faster, or even slower, than running Classic or HPC on CPU hardware. If running on a GPU device, high end NVidia graphics are strongly recommended. The performance of different NVidia cards varies by orders of magnitude – for benchmark tests using the original TUFLOW GPU solver review the <font color="blue"><tt>Hardware Benchmarking Wiki page</tt></font>. <br>
| |||