How can we help you?

IMPETUS Afea Solver is currently supported on the following 64-bit Operating Systems:

• Windows 10
• Redhat / CentOS 7
• Ubuntu 18.04 LTS

From version 6, IMPETUS Afea Solver requires CUDA 11 in order to utilize GPU acceleration. This means that the server needs to have CUDA 11 compatible drivers installed.

On 14th January 2020, Windows 7 and Windows Server 2012 reached the end of their life cycles and are no longer supported by Microsoft. Due to this, the latest Nvidia drivers no longer support these operating systems.

Our focus is to ensure a functional and a quality assured software that is reliable to all of our customers.

Each released version of IMPETUS Afea Solver’s computational modules (FE and DP) is subjected to version control based on a benchmark database. We test all our input commands to verify that they work correctly. For some applications, Recommended Modelling Practices (RMP) have been developed. The RMPs contain validation tests in combination with a recommended modelling methodology.

Our documents are updated based on the latest version control.

Yes, IMPETUS Afea Solver supports GPU acceleration utilizing Nvidia CUDA library. This means that we only support Nvidia graphics card (with CUDA support). Our software is built with GPU in mind and most parts of the software can utilize GPU acceleration. We currently support Mulit-GPU on a single node.

More information about Nvidia CUDA enabled applications can be found here:
GPU-ACCELERATED APPLICATIONS

IMPETUS Afea Solver currently supports CUDA-enabled devices with “compute capability” 3.5 or higher.

Please visit the link below to see if your graphics card is supported or not:
CUDA GPUs

While GPU acceleration usually speeds up a simulation, it’s actually slower when running small models. This is related to the GPU architecture, which is less efficient whenever only a small fraction of the available threads can be utilized.

Yes it is possible. To resume a simulation, use the “Resume simulation” button. This button will open a new dialog window with some options. The most simple way is to resume the simulation without any modifications. It’s also possible to increase the termination time or for more advanced options, one can modify termination time and the output frequncy.

(Be aware that resuming simulations might not work across different solver versions)

IMPETUS Afea Solver GUI is currently supported on the following 64-bit Operation Systems:

• Windows 10
• Windows 7
• Redhat / CentOS 7
• Ubuntu 18.04 LTS

We highly recommend using our built-in client-server interface. This allows to user to submit simulations and post process over the network. Read more about it in our
getting started document and our blog post about compute nodes.

Post processing over the network might not always be the most ideal way. If the simulation output is too large, the amount of data that is transferred between the client and the server might be too large and slow down the process. It that is the case, then we recommend using Windows Remote Desktop, VNC or TeamViewer (ask your IT department before using these since they might be blacklisted).

For more information about how to use IMPETUS Afea Solver with Microsoft Remote Desktop, please visit Running IMPETUS Afea Solver GUI remotely section here.

Port 8860 is the only one that needs to be open.

This problem can occur if there are jobs in the list that are/were executed on a network drive. If this network drive is not accessible through your client machine (e.g. credential signins are incorrect or not set), then this problem could occur. Make sure all network drives that are being used are properly configured.

This problem occurs if your display settings is set to scale text, apps and other items. The most likely reason is that your display is a High-DPI monitor and therefore scaling is enabled. To make the buttons visible, just resize the window vertically. If done correctly, the bottom buttons should now be visible.

This problem could occur if your multi-GPU system based on GeForce cards with SLI bridge enabled. Try to disable SLI in the Nvidia Control Panel.

This error message appears if the CUDA driver is insufficient to the CUDA runtime library. The most likely reason is that the graphics driver is too old. Download and update to the latest version and try again.

From version 6, IMPETUS Afea Solver requires CUDA 11 in order to utilize GPU acceleration. This means that the server needs to have CUDA 11 compatible drivers installed.

On 14th January 2020, Windows 7 and Windows Server 2012 reached the end of their life cycles and are no longer supported by Microsoft. Due to this, the latest Nvidia drivers no longer support these operating systems.

Latest Nvidia drivers can be found here:
NVIDIA Drivers

IMPETUS Afea Solver only supports Nvidia graphics cards as it utilizes CUDA. Minimum requirements is a graphics card with compute capability 3.5 (or higher).

It seems like one CUDA kernel or function (GPU) failed to execute. Please, contact our support with the error message. The error message is also written to gpu.info file. To make the process easier, please attach the gpu.info file to the email.
support@impetus.no

This error message appears if one part of the GPU code takes more than Y seconds to execute. As a safety mechanism, the OS stops the GPU in order to prevent the screen from freezing.

If you are running Linux, the reason might be that the GPU is used for display and running X (window manager). Try using a separate GPU for display and use your CUDA-enabled GPU as accelerator.

If you are running Windows, there are two possible reasons why this error message appears. You are either using a GeForce card as CUDA accelerator or a Tesla card with TCC turned off. If it’s the first case then the only possible way to solve this is to increase the watchdog timer. If it’s the second case then try enabling TCC mode for your Tesla card.

Increase watchdog timer

• Windows

The following steps describe how to enable TCC mode:

1. Open the command prompt (dos or powershell) as administrator
2. Navigate to “C:\Program Files\NVIDIA Corporation\NVSMI”
3. Type: “nvidia-smi.exe -dm 1”
4. Reboot the machine

This could come if multiple versions are installed. Most likely the software has been installed for the current user and for all users. You can try to fix this by running the Diagnostics tool.

To run the diagonstics tool, click on File -> Help -> Diagnostics
Then click on “Fix problems” button on top right corner.

The most likely scenario is that the license unit is not set for the current compute node. Open the License information window (the key-icon at the top right corner) and then click the button in the license unit box. Now, choose the compute node you wish to assign the unit to and then confirm by clicking the OK button.

If a unit is already assigned to the compute node, try to remove it and re-add it again and the click the OK button.