# Release Notes Archive
<a name="archive-detail-x86-gpu-pytorch-2.8-ubuntu-24-04"></a>

##### Release Date: 2025-08-14
<a name="2025-08-14-x86-gpu-pytorch-2.8-ubuntu-24-04"></a>

**AMI name**: Deep Learning OSS Nvidia Driver AMI GPU PyTorch 2.8 (Ubuntu 22.04) 20250814

##### Added
<a name="w2aab8c17b5b5c19c19b5b3b5"></a>
+ Added support for P5.4xlarge instances

##### Updated
<a name="w2aab8c17b5b5c19c19b5b3b7"></a>
+ Upgraded EFA to 1.43.1

##### Release Date: 2025-06-03
<a name="2025-06-03-x86-gpu-pytorch-2.8-ubuntu-24-04"></a>

**AMI name**: Deep Learning OSS Nvidia Driver AMI GPU PyTorch 2.8 (Ubuntu 22.04) 20250602

##### Added
<a name="w2aab8c17b5b5c19c19b5b5b5"></a>
+ Initial release of Deep Learning AMI GPU PyTorch 2.8 (Ubuntu 22.04) series. Including a Python virtual environment pytorch (source /opt/pytorch/bin/activate) complimented with NVIDIA Driver R570, CUDA=12.8, cuDNN=9.10, PyTorch NCCL=2.26.5, and EFA=1.40.0.

##### Known Issues
<a name="w2aab8c17b5b5c19c19b5b5b7"></a>
+ "With compute capability sm10.0 (Blackwell-architecture) GPUs, the FP8 datatype with scaled dot-product attention contains a deadlock that causes the kernel to hang under some circumstances, such as when the problem size is large or the GPU is running multiple kernels simultaneously. A fix is planned for a future release." [[cuDNN 9.10.0 release notes](https://docs.nvidia.com/deeplearning/cudnn/backend/latest/release-notes.html#cudnn-9-10-0)]
  + For users seeking to run P6-B200 instances with FP8 data and scaled dot-product attention, please consider installing flash attention manually.