cuDNN 8.2.2.26-CUDA-11.4.1
There is a newer install of cuDNN
The NVIDIA CUDA Deep Neural Network library (cuDNN) is a GPU-accelerated library of primitives for deep neural networks. cuDNN provides highly tuned implementations for standard routines such as forward and backward convolution, pooling, normalization, and activation layers. cuDNN is part of the NVIDIA Deep Learning SDK.Accessing cuDNN 8.2.2.26-CUDA-11.4.1
To load the module for cuDNN 8.2.2.26-CUDA-11.4.1 please use this command on the BEAR systems (BlueBEAR and BEAR Cloud VMs):
📋
module load bear-apps/2021b
module load cuDNN/8.2.2.26-CUDA-11.4.1
BEAR Apps Version
Architectures
EL8-icelake — EL8-sapphirerapids
The listed architectures consist of two part: OS-CPU. The OS used is represented by EL and there are several different processor (CPU) types available on BlueBEAR. More information about the processor types on BlueBEAR is available on the BlueBEAR Job Submission page.
More Information
For more information visit the cuDNN website.
Dependencies
This version of cuDNN has a direct dependency on: CUDA/11.4.1
Required By
This version of cuDNN is a direct dependent of: CuPy/11.4.0-foss-2021b-CUDA-11.4.1 jax/0.3.13-foss-2021b-CUDA-11.4.1 TensorFlow/2.7.1-foss-2021b-CUDA-11.4.1
Other Versions
These versions of cuDNN are available on the BEAR systems (BlueBEAR and BEAR Cloud VMs). These will be retained in accordance with our Applications Support and Retention Policy.
| Version | BEAR Apps Version |
|---|---|
| 8.4.1.50-CUDA-11.7.0 | 2022a |
| 8.2.1.32-CUDA-11.3.1 | 2021a |
| 8.0.4.30-CUDA-11.1.1 | 2020b |
| 7.6.4.38-gcccuda-2019b | 2019b |
Last modified on 24th June 2022