The vulnerability of the Checkpointing module in the NVIDIA Resiliency Extension (NVRx) toolset, which is designed to enhance the fault tolerance of large-scale distributed training systems. This vulnerability allows attackers to gain unauthorized access to protected information, replace data, enhance their privileges, or trigger service failures.

The vulnerability of the Checkpointing module in the NVIDIA Resiliency Extension NVRx toolset, which is designed to improve the fault tolerance of large-scale distributed training systems, is related to synchronization errors when using a shared resource. Exploiting this vulnerability can allow...