Frameworks Support Matrix

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9²

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9³

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁴

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁵

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁶

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁷

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+^{https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html#fntarg_1}

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁸

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9⁹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

Release 24.09 is based on CUDA 12.6.1 which requires NVIDIA Driver release 560 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, R520, R530, R545 and R555 drivers, which are not forward-compatible with CUDA 12.6. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

.

Release 24.08 is based on CUDA 12.6 which requires NVIDIA Driver release 560 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, R520, R530, R545 and R555 drivers, which are not forward-compatible with CUDA 12.6. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

.

Release 24.07 is based on CUDA 12.5.1 which requires NVIDIA Driver release 555 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, R520 and R545 drivers, which are not forward-compatible with CUDA 12.5. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.06 is based on CUDA 12.4.1, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.05 is based on CUDA 12.4.1, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.04 is based on CUDA 12.4.1, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.03 is based on CUDA 12.4.0.41, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.02 is based on CUDA 12.3.2, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 24.01 is based on CUDA 12.3.2, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 470.57 (or later R470), 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R450, R460, R510, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• OpenMPI 4.1.7
• OpenUCX 1.15.0

• OpenMPI 4.1.7
• OpenUCX 1.15.0

• OpenMPI 4.1.7
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• RAPIDS 24.08
• PyTorch 2.5.0a0+b465a5843b
• Apex
• GDRCopy 2.3
• TensorBoard 2.12.0
• Torch-TensorRT 2.5.0.a0
• MAGMA 2.6.2
• JupyterLab 4.2.5 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.10
• NVSHMEM 2.10.1

• RAPIDS 24.04
• PyTorch 2.4.0a0+3bcc3cddb5
• Apex
• GDRCopy 2.3
• TensorBoard 2.12.0
• Torch-TensorRT 2.4.0.a0
• MAGMA 2.6.2
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.8
• NVSHMEM 2.10.1

• RAPIDS 24.04
• PyTorch 2.4.0a0+07cecf4
• Apex
• GDRCopy 2.3
• TensorBoard 2.12.0
• Torch-TensorRT 2.4.0.a0
• MAGMA 2.6.2
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.6
• NVSHMEM 2.10.1

• RAPIDS 24.02
• PyTorch 2.3.0a0+6ddf5cf85e
• Apex
• GDRCopy 2.3
• TensorBoard 2.12.0
• Torch-TensorRT 2.30.a0
• MAGMA 2.6.2
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.5
• NVSHMEM 2.10.1

• PyTorch 2.3.0a0+40ec155e58
• APEX
• MAGMA 2.6.2
• TensorBoard 2.12.0
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.4

• PyTorch 2.2.0a0+81ea7a4

• APEX
• MAGMA 2.6.2
• TensorBoard 2.12.0
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.2.1

• Ubuntu 22.04.2 LTS with Python 3.10
• NVIDIA cuDNN 9.1.0.70
• NVIDIA NCCL 2.21.5
• NVIDIA HPC-X 2.18
• Nsight Compute 2024.1.1.0
• Nsight Systems 2024.2.1.106
• TransformerEngine 1.5

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Python 3.10
• OpenMPI 4.1.7
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.7
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.7
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 3.0.2

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.5.0a0
• MAGMA 2.6.2
• RAPIDS 24.08
• JupyterLab 4.2.5 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.10

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.4.0a0
• MAGMA 2.6.2
• RAPIDS 24.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.8

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.4.0a0
• MAGMA 2.6.2
• RAPIDS 24.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.6

• Python 3.10
• APEX
• TensorBoard 2.16.2
• Torch-TensorRT 2.5.0a0
• MAGMA 2.6.2
• RAPIDS 24.06
• JupyterLab 4.2.5 including Jupyter-TensorBoard
• TensorRT Model Optimizer 0.15.1
• TransformerEngine v1.10

• Python 3.10
• APEX
• TensorBoard 2.16.2
• Torch-TensorRT 2.5.0a0
• MAGMA 2.6.2
• RAPIDS 24.06
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• TensorRT Model Optimizer 0.15.0
• TransformerEngine v1.9

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.4.0a0
• MAGMA 2.6.2
• RAPIDS 24.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.8

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.4.0a0
• MAGMA 2.6.2
• RAPIDS 24.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.7

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.4.0a0
• MAGMA 2.6.2
• RAPIDS 24.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.6

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.3.0a0
• MAGMA 2.6.2
• RAPIDS 24.02
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.5

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 24.02
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.4

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.12
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.2.1

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.12
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine v1.2.1

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.08
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 4.2.5
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 24.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.12
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.12
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.47 including

• PyTorch2.5.0a0+b465a5843b
• TensorFlow 2.16.0
• NVIDIA CUDA 12.6.1
• NVIDIA cuBLAS 12.6.3.1
• cuDNN 9.4.0.58
• NVIDIA NCCL 2.22.3 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 10.4.0.26
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.20
• OpenMPI 4.1.7
• FIL
• NVIDIA DALI® 1.41
• ONNX Runtime 1.19.2
• Intel OpenVINO 2024.0.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.13.0
• vLLM version 0.5.3 post1
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.47 including

• PyTorch2.5.0a0+872d972e41
• TensorFlow 2.16.0
• NVIDIA CUDA 12.6
• NVIDIA cuBLAS 12.6.0.22
• cuDNN 9.3.0.75
• NVIDIA NCCL 2.22.3 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 10.3.0.26
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.19
• OpenMPI 4.1.7
• FIL
• NVIDIA DALI® 1.40
• ONNX Runtime 1.18.1
• Intel OpenVINO 2024.0.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.12.0
• vLLM version 0.5.3 post1
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.47 including

• PyTorch2.4.0a0+3bcc3cddb5
• TensorFlow 2.16.0
• NVIDIA CUDA 12.5.1
• NVIDIA cuBLAS 12.5.3.2
• cuDNN 9.2.1.18
• NVIDIA NCCL 2.22.3 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 10.2.0.19
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.19
• OpenMPI 4.1.7
• FIL
• NVIDIA DALI® 1.39
• ONNX Runtime 1.18.1
• Intel OpenVINO 2024.0.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.11.0
• vLLM version 0.5.0 post1
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.47 including

• PyTorch 2.4.0a0+f70bf71
• TensorFlow 2.15.0
• NVIDIA CUDA 12.5.0.23
• NVIDIA cuBLAS 12.5.2.13
• cuDNN 9.1.0.70
• NVIDIA NCCL 2.21.5 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 10.1.0.27
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.19
• OpenMPI 4.1.4+
• FIL
• NVIDIA DALI® 1.38
• ONNX Runtime 1.18.0
• Intel OpenVINO 2024.0.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.10.0
• vLLM version 0.4.3
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.46 including

• PyTorch 2.4.0a0+07cecf4
• TensorFlow 2.15.0
• NVIDIA CUDA 12.4.1
• NVIDIA cuBLAS 12.4.5.8
• cuDNN 9.1.0.70
• NVIDIA NCCL 2.21.5 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 10.0.1.6
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.19
• OpenMPI 4.1.4+
• FIL
• NVIDIA DALI® 1.37
• ONNX Runtime 1.18.0
• Intel OpenVINO 2023.3.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.9
• vLLM version 0.4.0 post1
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.45 including

• PyTorch 2.3.0a0+6ddf5cf85e
• TensorFlow 2.15.0
• NVIDIA CUDA 12.4.1
• NVIDIA cuBLAS 12.4.5.8
• cuDNN 9.1.0.70
• NVIDIA NCCL 2.21.5 (optimized for NVIDIA NVLink®)
• Amazon Labs Sockeye sequence-to-sequence framework 2.3.14 (for machine translation)
• NVIDIA TensorRT™ 8.6.3
• OpenUCX 1.15.0
• GDRCopy 2.3
• NVIDIA HPC-X 2.18
• OpenMPI 4.1.4+
• FIL
• NVIDIA DALI® 1.36
• ONNX Runtime 1.17.3
• Intel OpenVINO 2023.3.0
• DCGM 3.2.6
• TensorRT-LLM version release/0.9
• vLLM version 0.4.0 post1
• nvImageCodec 0.2.0.7

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.43 including

• PyTorch 2.3.0a0+40ec155e58
• TensorFlow 2.15.0
• Intel OpenVINO 2023.3.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.8
• vLLM 0.3.0

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.43 including

• PyTorch 2.3.0a0+ebedce2
• TensorFlow 2.15.0
• Intel OpenVINO 2023.3.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.8
• vLLM 0.3.0

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.41 including

• PyTorch 2.2.0a0+81ea7a4
• TensorFlow 2.12.0
• ONNX Runtime 1.16.3
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.7.1
• vLLM 0.2.3

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹⁰

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL9¹

Red Hat Enterprise Linux 8 / CentOS 8¹ (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 9 / CentOS 9¹

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+¹² (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+
• 6.0

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7¹

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

Release 23.12 is based on CUDA 12.3.2, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R460, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

Release 23.11 is based on CUDA 12.3.0, which requires NVIDIA Driver release 545 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 525.85 (or later R525), 535.86 (or later R535), or 545.23 (or later R545).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R460, and R520 drivers, which are not forward-compatible with CUDA 12.3. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.2. ¹³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.2. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.2. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.1. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.1. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.1. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.0. ³

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.0. ³

Release 23.02 is based on CUDA 12.0.1, which requires NVIDIA Driver release 525 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), 515.65 (or later R515), or 525.85 (or later R525).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.0. ³

Release 23.01 is based on CUDA 12.0.1, which requires NVIDIA Driver release 525 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), 515.65 (or later R515), or 525.85 (or later R525).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 12.0. ³

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal
  Note:

  Starting with the 23.06 release, containers will no longer be tested on Pascal GPU architecture.

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.15.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• PyTorch 23.11

• APEX
• MAGMA 2.6.2
• TensorBoard 2.12.0
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12

• PyTorch 23.09

• APEX
• MAGMA 2.6.2
• TensorBoard 2.12.0
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12

• PyTorch 23.07

• APEX
• MAGMA 2.6.2
• TensorBoard 2.12.0
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.10.0+96ed6fc

• Python 3.10.12
• Nsight Compute 2023.2.2.3
• Nsight Systems 2023.3.1.92
• NVIDIA CUDA 12.2.0
• NVIDIA DALI 1.30.0 in T5x containers
• Mellanox OFED 5.9-0.5.6.0
• Transformer Engine
  • 1.0.0.dev0+1f4c397 in Paxml container
  • 0.13.0+511241c in T5X container

• Python 3.10.12
• Nsight Compute 2023.2.0
• Nsight Systems 2023.2.0
• NVIDIA CUDA 12.1.1
• NVIDIA DALI 1.28.0 in Paxml and T5x containers
• Mellanox OFED 5.9-0.5.6.0
• Transformer Engine 0.13.0.dev0+03202c3

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.10

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.28 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.10
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.8
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.8
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.8
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.4
• SHARP 2.6.0

• Python 3.8
• OpenMPI 4.1.4+
• GDRCopy 2.3
• Paddle-TRT 2.3.2
• SHARP 2.6.0

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.10
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12+170797

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.10
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12+170797

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.08
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12+170797

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.08
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.12

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 2.0.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.06
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.11.0++3f01b4f

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.5.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.06
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.10.0+96ed6fc

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.5.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.9

• Python 3.10
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.4.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.04
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.8

• Python 3.8
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.4.0.dev0
• MAGMA 2.6.2
• RAPIDS 23.02
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.7

• Python 3.8
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 23.02
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.6.0

• Python 3.8
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.12.1
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.5.0

• Python 3.8
• APEX
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.12.1
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• PyTorch quantization wheel v2.1.2
• TransformerEngine 0.3.0

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.10
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.10
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.08
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.4.10
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.08
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.13.0
• RAPIDS 23.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.1
• TensorBoard 2.12.0
• RAPIDS 23.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.28.0
• TensorBoard 2.12.0
• RAPIDS 23.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.10
• Horovod 0.27.0
• TensorBoard 2.12.0
• RAPIDS 23.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.27.0
• TensorBoard 2.12.0
• RAPIDS 23.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.27.0
• TensorBoard 2.9.0
• RAPIDS 23.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.27.0
• TensorBoard 1.15.5
• RAPIDS 23.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.26.1
• TensorBoard 2.9.0
• RAPIDS 22.12.0
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.26.1
• TensorBoard 1.15.5
• RAPIDS 22.12.0
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.12.0
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• RAPIDS 22.12.0
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.41 including

• PyTorch 2.2.0a0+81ea7a48
• TensorFlow 2.12.0
• ONNX Runtime 1.16.3
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.7.0
• vLLM 0.2.3

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.40 including

• PyTorch 2.2.0a0+6a974be
• TensorFlow 2.12.0
• ONNX Runtime 1.16.3
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.6.0
• vLLM 0.2.0

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.39 including

• PyTorch 2.1.0a0+32f93b1
• TensorFlow 2.12.0
• ONNX Runtime 1.16.0
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL
• TensorRT-LLM release/0.5.0
• vLLM 0.2.0

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.38 including

• PyTorch 2.1.0a0+32f93b1
• TensorFlow 2.12.0
• ONNX Runtime 1.16.0
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.37 including

• PyTorch 2.1.0a0+29c30b1
• TensorFlow 2.12.0
• ONNX Runtime 1.15.1
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.36 including

• PyTorch 2.1.0a0+b5021ba
• TensorFlow 2.12.0
• ONNX Runtime 1.15.0
• Intel OpenVINO 2023.0.0
• DCGM 2.4.7
• FIL

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.35 including

• PyTorch 2.1.0a0+4136153
• TensorFlow 2.12.0
• ONNX Runtime 1.15.0
• Intel OpenVINO 2022.1.0
• DCGM 2.4.7
• FIL

• AMD x86 CPU
• Intel x86 CPU
• Python 3.10

2.34 including

• PyTorch 2.0.0
• TensorFlow 2.12.0
• ONNX Runtime 1.15.0
• Intel OpenVINO 2022.1.0
• DCGM 2.4.7
• FIL

• AMD x86 CPU
• Intel x86 CPU

2.33 including

• PyTorch 2.1.0a0+fe05266f
• TensorFlow 2.12.0
• ONNX Runtime 1.14.1
• Intel OpenVINO 2022.1.0
• DCGM 2.2.9
• FIL

• AMD x86 CPU
• Intel x86 CPU

2.32 including

• PyTorch 2.0.0a0+1767026
• TensorFlow 2.11.0
• ONNX Runtime 1.14.1
• Intel OpenVINO 2022.1.0
• DCGM 2.2.9
• FIL

• AMD x86 CPU
• Intel x86 CPU

2.31 including

• PyTorch 1.14.0a0+44dac51
• TensorFlow 1.15.5
• TensorFlow 2.11.0
• ONNX Runtime 1.13.1
• Intel OpenVINO 2022.1.0
• DCGM 2.2.9
• FIL

• AMD x86 CPU
• Intel x86 CPU

2.30 including

• PyTorch 1.14.0a0+44dac51
• TensorFlow 1.15.5
• TensorFlow 2.11.0
• ONNX Runtime 1.13.1
• Intel OpenVINO 2022.1.0
• DCGM 2.2.9
• FIL

• AGX Xavier

• AGX Xavier

• AGX Xavier

• AGX Xavier

• AGX Xavier

• AGX Xavier

• AGX Xavier

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100
• NVIDIA DGX H100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100

• NVIDIA DGX-1™
• DGX-2™
• NVIDIA DGX™ A100
• DGX Station
• DGX Station A100

• DGX-1
• DGX-2
• NVIDIA DGX A100
• DGX Station
• DGX Station A100

• DGX-1
• DGX-2
• DGX A100
• DGX Station
• DGX Station A100

• DGX-1
• DGX-2
• DGX A100
• DGX Station
• DGX Station A100

• DGX-1
• DGX-2
• DGX A100
• DGX Station
• DGX Station A100

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+1 (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0+

Red Hat Enterprise Linux 7 / CentOS 72

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

• 4.1+¹ (4.6+ multi-node NCCL)
• 4.99.x (DGX A100)
• 5.0
• 5.1

Red Hat Enterprise Linux 7 / CentOS 7²

• EL7-20.02+¹

Red Hat Enterprise Linux 8 / CentOS 8² (All DGX systems except DGX Station A100)

• EL8-20.11+¹

Release 22.12 is based on CUDA 11.8.0, which requires NVIDIA Driver release 520 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 515.65 (or later R515).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.8. ³

Release 22.11 is based on CUDA 11.8.0, which requires NVIDIA Driver release 520 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 515.65 (or later R515).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.8. ³

Release 22.10 is based on CUDA 11.8.0, which requires NVIDIA Driver release 520 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 515.65 (or later R515).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.8. ³

Release 22.09 is based on CUDA 11.8.0, which requires NVIDIA Driver release 520 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), or 515.65 (or later R515).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.8. ³

Release 22.08 is based on CUDA 11.7.1, which requires NVIDIA Driver release 515 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), or 510.47 (or later R510).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.8. ³

Release 22.07 is based on CUDA 11.7 Update 1 Preview, which requires NVIDIA Driver release 515 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), or 510.47 (or later R510).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.7. ³

Release 22.06 is based on CUDA 11.7 Update 1 Preview, which requires NVIDIA Driver release 515 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), or 510.47 (or later R510).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.7. ³

Release 22.05 is based on CUDA 11.7, which requires NVIDIA Driver release 515 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), or 510.47 (or later R510).

The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, and R460 drivers, which are not forward-compatible with CUDA 11.7. ³

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ada Lovelace Architecture
• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Hopper™ Architecture
• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• NVIDIA Ampere GPU Architecture
• NVIDIA Turing™
• NVIDIA Volta™
• NVIDIA Pascal

• Ampere GPU Architecture
• Turing
• Volta
• Pascal

• NVIDIA Ampere GPU Architecture
• Turing
• Volta
• Pascal

• NVIDIA Ampere GPU Architecture
• Turing
• Volta
• Pascal

• NVIDIA Ampere GPU Architecture
• Turing
• Volta
• Pascal

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.4+
• OpenUCX 1.14.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• OpenMPI 4.1.2rc4+
• OpenUCX 1.12.0

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• 26b9f648
• Python 3.8

• rdma-core 32.1
• Nsight Compute 2022.3.0.22
• Horovod 0.26.1 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Nsight Compute 2022.3.0.22
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Nsight Compute 2022.3.0.22
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Nsight Compute 2022.3.0.22
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• JupyterLab 2.2.10 with updated mistune 2.0.4

• rdma-core 32.1
• Nsight Compute 2022.2.1.0
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• rdma-core 32.1
• Nsight Compute 2022.2.1.0
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• rdma-core 32.1
• Nsight Compute 2022.1.1.2
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• rdma-core 32.1
• Nsight Compute 2022.1.1.2
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• rdma-core 32.1
• OpenMPI 4.1.1rc1
• OpenUCX 1.12
• NVIDIA HPC-X 2.10
• Nsight Compute 2022.1.1.2
• Nsight Systems 2022.2.1.31-5fe97ab
• Horovod 0.24.2 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• TensorRT 8.2.4.2
• DALI 1.12.0
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• rdma-core 32.1
• OpenMPI 4.1.1rc1
• OpenUCX 1.12
• NVIDIA HPC-X 2.10
• Nsight Compute 2022.1.1.2
• Nsight Systems 2021.2.5.2.53
• Horovod 0.23.0 with enhancements
• GluonCV Toolkit 0.7
• GluonNLP Toolkit 0.8.1
• TensorRT 8.2.3
• DALI 1.11.1
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.2.10
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.3.0.0
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.10.01
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• TransformerEngine 0.3.0

• Python 3.8
• APEX
• Nsight Compute 2022.3.0.0
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.10
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• TransformerEngine 0.2.0

• Python 3.8
• APEX
• Nsight Compute 2022.3.0.22
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.08.01
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• TransformerEngine 0.1.0

• Python 3.8
• APEX
• Nsight Compute 2022.3.0.22
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.08.01
• JupyterLab 2.3.2 including Jupyter-TensorBoard
• TransformerEngine 0.1.0

• Python 3.8
• APEX
• Nsight Compute 2022.2.1
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.2.1
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.2.1
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.2.0.13
• TensorBoard 2.9.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.6.2
• RAPIDS 22.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.1.1.2
• TensorBoard 2.8.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.5.2
• RAPIDS 22.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.1.1.2
• TensorBoard 2.8.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.5.2
• RAPIDS 22.02
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.1.0.0
• TensorBoard 2.8.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.5.2
• RAPIDS 21.12
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• APEX
• Nsight Compute 2022.1.0.0
• TensorBoard 2.8.0
• Torch-TensorRT 1.1.0a0
• MAGMA 2.5.2
• RAPIDS 21.12
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.6
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.10.01
• Nsight Compute 2022.3.0.22
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.3.0.22
• RAPIDS 22.10.01
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.10
• Nsight Compute 2022.3.0.22
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.3.0.22
• RAPIDS 22.10
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.08.01
• Nsight Compute 2022.3.0.22
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.3.0.22
• RAPIDS 22.08.01
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.08.01
• Nsight Compute 2022.3.0.22
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.3.0.22
• RAPIDS 22.08.01
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.06
• Nsight Compute 2022.2.1.3
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.2.1.3
• RAPIDS 22.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.06
• Nsight Compute 2022.2.1.3
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.2.1.3
• RAPIDS 22.06
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 2.9.0
• RAPIDS 22.04
• Nsight Compute 2022.2.1.3
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 5.7.8
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

• Python 3.8
• Horovod 0.24.3
• TensorBoard 1.15.5
• Nsight Compute 2022.2.1.3
• RAPIDS 22.04
• Jupyter and JupyterLab:
  • Jupyter Client 6.0.0
  • Jupyter Core 4.6.1
  • Jupyter Notebook 6.0.3
  • JupyterLab 2.3.2
  • JupyterLab Server 1.0.7
  • Jupyter-TensorBoard