Your system must meet the following requirements to use the GPU accelerator capability in Mechanical APDL:
The machine(s) being used for the simulation must contain at least one supported NVIDIA GPU card or one Intel Xeon Phi coprocessor. The following cards are supported:
| NVIDIA Tesla Series (any model) |
| NVIDIA Quadro GP100 |
| NVIDIA Quadro P6000 |
| NVIDIA Quadro P5000 |
| NVIDIA Quadro M6000 |
| NVIDIA Quadro K6000 |
| NVIDIA Quadro K5200 |
| NVIDIA Quadro K5000 |
| Intel Xeon Phi 7120 |
| Intel Xeon Phi 5110 |
| Intel Xeon Phi 3120 |
When using the sparse solver or eigensolvers based on the sparse solver (for example, Block Lanczos or subspace), only nVIDIA GPU devices with significant double precision performance are recommended in order to achieve optimal performance. These include the following models:
| NVIDIA Tesla Series (any model except the following: K10/M4/M40/M6/M60/P4/P40) |
| NVIDIA Quadro GP100 |
| NVIDIA Quadro K6000 |
For NVIDIA GPU cards, the driver version must be 375.20 or newer.
For Intel Xeon Phi coprocessors, the driver version must be “MPSS 3.8” or newer.
Note: Mechanical APDL assumes that the Intel Xeon Phi coprocessor driver is installed in
/opt/intel/mic/coi. If the installation is located in a different folder, set the
ANS_MIC_LIB_PATH environment variable to the path where the Xeon
Phi driver is installed (for example,
/mypath/mic/coi).