Frequently Asked Questions

These Frequently Asked Questions are designed to provide an easy understanding of general questions aswell as problems that could occur with our softwares.

IMPETUS Afea Solver Engine

Which platforms are supported?

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

  • Windows XP
  • Windows Vista
  • Windows 7
  • Windows 10
  • Redhat / CentOS 6
  • Ubuntu 16

We highly recommend a Windows platform since it’s provided with a graphical user interface for easier use.

How reliable are the results generated by IMPETUS Afea Solver?

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.

Documents on our support page are updated based on the latest version control.

Does IMPETUS Afea Solver support GPU acceleration?

Yes, IMPETUS Afea Solver supports GPU acceleration utilizing Nvidia CUDA library. We currently support Mulit-GPU on a single node.

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

Which graphics cards are supported?

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

I ran a small model twice, one with and one without GPU acceleration. The GPU simulation was slower than the CPU simulation. How come?

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.

Is it possible to resume a simulation?

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

I have a High-DPI monitor and Windows 10. When I open the IMPETUS Afea Solver GUI the text is too large and the icons are too small. How can I make it look good?

If you have a High-DPI monitor, you may have a very large resolution and automatic scaling done by Windows. For instance you may have a resolution of 3200×1600 and text scaled to 200%. This does not work well with IMPETUS Afea Solver GUI. We recommend to set the scaling 100% and instead reduce the resolution if necessary.

To do this, right click on the desktop background, select Display settings, and set the slider “Change the size of text, apps and other items:” to 100%. If text and icons are too small, reduce the resolution to for instance 2048×1152 or 1920×1080, depending on the aspect ratio of the monitor.

Which platforms are supported?

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

  • Windows XP
  • Windows Vista
  • Windows 7
  • Redhat / CentOS 6
  • Ubuntu 16

How do I connect to the server over the network?

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.

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.

Troubleshooting

When I start the installer, a window pops up but there are no buttons to proceed, what should I do?

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.

I’m having problems with the 3D graphics on my laptop, I can’t create screenshots or videos. What is wrong?

If the laptop has an NVIDIA graphics processor, make sure it is used and not the integrated graphics processor. Open NVIDIA Control Panel and go to Manage 3D Settings. Under Preferred graphics processor, if there is a choice between Auto, High-performance NVIDIA processor and Integrated graphics, select High-performance NVIDIA processor and Apply the changes.

I’m running on a multi-GPU system and when I start a simulation on one card, the same amount of memory is allocated on the other card. What’s wrong?

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

I get the error message: “This software requires a newer version of the graphics driver”. What is wrong?

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.

Latest Nvidia drivers can be found here:
NVIDIA Drivers

I get the error message: “Error: No CUDA 2.0 (or higher) capable devices found”. What is wrong?

IMPETUS Afea Solver only supports Nvidia graphics cards as it utilizes CUDA. Minimum requirements on Windows is a graphics card with compute capability 2.0 while Linux versions requires 3.5 (or higher). Support of compute capability 2.0 on Windows is depricated and will be removed in future releases.

I get the error message: “CUDA error: XXXX in file YYYY, line ZZZZ”. What is wrong?

It seems like one CUDA kernel or function (GPU) failed to execute. Please, contact us at support@impetus.no with the error message. The error message is also written to gpu.info file.

I get the error message: “the launch timed out and was terminated”. What is wrong?

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

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