rsnt_translations
56,420
edits
VASP: Difference between revisions
Jump to navigation
Jump to search
no edit summary
No edit summary |
No edit summary |
||
| Line 62: | Line 62: | ||
== Vasp-GPU == <!--T:25--> | == Vasp-GPU == <!--T:25--> | ||
[[File:VASP benchmark ENG.png|400px|thumb| Fig.1 Simulation time as a function of the number of CPUs for GPU=0, 1, and 2]] | [[File:VASP benchmark ENG.png|400px|thumb| Fig.1 Simulation time as a function of the number of CPUs for GPU=0, 1, and 2]] | ||
Vasp-GPU executable files run on both GPUs and CPUs of a node. Basically, calculation on a GPU is much more expensive than on a CPU, therefore we highly recommend to perform a benchmark using one or 2 GPUs to make sure they are getting a maximum performance from the GPU use. Fig.1 shows a benchmark of Si crystal which contains 256 Si-atoms in the | Vasp-GPU executable files run on both GPUs and CPUs of a node. Basically, calculation on a GPU is much more expensive than on a CPU, therefore we highly recommend to perform a benchmark using one or 2 GPUs to make sure they are getting a maximum performance from the GPU use. Fig.1 shows a benchmark of Si crystal which contains 256 Si-atoms in the simulation box. Blue, black and red lines show simulation time as a function of Number of CPU for GPU=0, 1, and 2 respectively. It shows the performance for GPU=1,2 and CPU=1 is more than 5 times better compared to GPU=0 and CPU=1. However, a comparison of calculations with GPU=1 and GPU=2 indicates that there is not much performance gain from GPU=1 to GPU=2. In fact, use for GPU=2 is around 50% in our monitoring system. Therefore we recommend users to first perform a benchmark like this for their own system to make sure they are not wasting any computer resources. | ||
== Building VASP yourself == <!--T:12--> | == Building VASP yourself == <!--T:12--> | ||