Developed and integrated FP64 Tensor Core support for NVIDIA GPUs in the modular/modular repository, enabling double-precision tensor core operations to enhance numerical fidelity for precision-critical workloads. The work involved updating Mojo source files, including tensor_core.mojo and _mma_nvidia.mojo, and implementing a comprehensive validation suite to ensure correctness across NVIDIA platforms, particularly the GH200. Leveraged GPU programming and high-performance computing skills to design and test the new feature, utilizing Mojo and Bazel-based CI workflows. This contribution addressed a tracked issue, improved documentation, and positioned the codebase to support advanced scientific, simulation, and finance applications requiring high-precision GPU acceleration.