Good Hardware Means Improved FlowSight Performance

In order to take full advantage of FlowSight, our advanced state-of-the-art postprocessor, it is essential that you have good hardware. In this blog, Stephen Sanchez, senior GUI developer, gives his two cents on how you can obtain improved FlowSight performance by following these hardware recommendations.

Get a really good graphics card

We highly recommend that you start with a graphics card with at least 3GB of VRAM. This is especially important if you will be doing a lot of volume rendering. Volume rendering is an advanced capability of FlowSight that visualizes the details of a variable throughout the fluid domain, instead of just the iso-surface. This feature is quite insightful, but requires good hardware to be effectively used during post-processing.

Volume rendering FlowSight performance
Volume rendering (left) of bubbles in FlowSight.The image shows the strain rate magnitude of each bubble and the surrounding liquid.

Next, you should not use Intel integrated graphics as your primary graphics hardware. Much of FlowSight’s functionality does not work with this configuration, and as a result, we do not support Intel integrated graphics. FlowSight performs best when used with NVIDIA graphics cards, particularly the Quadro family. While high-end AMD cards should work, we have found that they are not as reliable as NVIDIA hardware and drivers, so we always recommend NVIDIA over AMD.

Nvidia graphics card

Dual graphics cards on laptops – A simple but hidden solution

Many laptops now come with the ability to switch between a NVIDIA card and an Intel Integrated graphics card. It is important that you make sure FlowSight (as well FLOW-3D) is being launched with the NVIDIA card. Forcing your laptop to launch with the NVIDIA card can be done through the NVIDIA control panel.

Switching graphics card to Nvidia

Update your video driver

We recommend that you check to make sure that your video driver is up-to-date. We have had reports of artifacts and display issues in FlowSight that have been easily resolved by simply updating the video driver. Keeping your video driver current is a good way to avoid such issues.

RAM, RAM, RAM!

It is important to be aware of memory requirements, as insufficient memory can lead to as high as a 10x performance decrease! The amount of RAM needed depends on a number of factors, particularly the size of your simulation. In order to provide users with the most flexibility, we have the following RAM recommendations based on simulation size:

  • Extra-large (200 million+ cells): At least 128GB
  • Large (Between 60-150 million cells): 64-128GB
  • Medium (Between 30-60 million cells): 32-64GB
  • Small (30 million cells and below): At least 32GB

However, you should always get the most RAM as you can, irrespective of the problem size in order to maximize flexibility and ensure the smoothest user experience.

In a later blog we’ll expand on how to get the most out of FlowSight by discussing core features that are typically the most time and resource intensive, and how they can be used more efficiently.