web
You’re offline. This is a read only version of the page.
close
Thermo Fisher Scientific > Electron Microscopy > Electron Microscopes > 3D Visualization, Analysis and EM Software > Customer Portal > Support Center

VISUALIZATION AND DATA SOLUTIONS

CUSTOMER SUPPORT CENTER

Views:

Amira 3D, Avizo 3D

 

Issue

 

When using Deep Learning features such as AI Assisisted Segmentation or StarDist Segmentation, the feature cannot be used and an error message appears. The error details include the following message: "RuntimeError: CUDA error: no kernel image is available for execution on the device”.

 

Reason

 

Although all causes mentioned in the error message should be considered, when the details specifically include:

"RuntimeError: CUDA error: no kernel image is available for execution on the device”

it generally indicates that NVIDIA's CUDA system—used for AI, Deep Learning, and other advanced computations—is unable to detect the graphics card properly.

 

Prerequisites

 

  • You must have an appropriate license to use these features. Most of the affected features are included in Amira 3D Pro and Avizo 3D Pro editions.
  • Only NVIDIA graphics cards are supported for AI and Deep Learning features.

 

Suggested Solutions

 

1. Verify that your graphics card is supported by your current version of Amira-Avizo.

1.1. Amira-Avizo 2025.1 and later
Supports and requires an NVIDIA GPU with a compute capability of at least 7.5.

1.2. Amira-Avizo 2022.1 to 2024.2
Supports and requires an NVIDIA GPU with a compute capability

  • of 3.5 to 11.0 (inclusive) for 2D
  • of 5.2 to 11.0 (inclusive) for 3D.

1.3. Amira-Avizo 2019.3 to 2021.2
Supports and requires an NVIDIA GPU with a compute capability of 3.5 to 11.0 (inclusive).

2. Update your NVIDIA driver.
The drivers should be updated to the latest stable version.
Official NVIDIA drivers can be downloaded and installed using the NVIDIA App or directly from the NVIDIA website: https://www.nvidia.com/en-us/drivers/.

 

Related Links