{"enrichment":{"capability":"NVSHMEM provides a parallel programming interface for NVIDIA GPU clusters, creating a global address space across multiple GPUs' memory that supports fine-grained GPU-initiated and CPU-initiated communication operations.","verdict":"NVSHMEM is a recently released (27 days old), actively maintained GPU cluster communication library in Beta status, positioned for HPC and ML workloads. Medium install friction and unclear licensing are the primary concerns; no known vulnerabilities. Suitable for developers working with multi-GPU NVIDIA systems who can verify the license terms independently."},"id":"nvidia-nvshmem-cu13","links":{"html":"https://skillfed.io/packages/nvidia-nvshmem-cu13","md":"https://skillfed.io/packages/nvidia-nvshmem-cu13.md","pypi":"https://pypi.org/project/nvidia-nvshmem-cu13/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-07-17","license_spdx":null,"license_treatment":"unclear","name":"nvidia-nvshmem-cu13","python_support":"supports_current","summary":"NVSHMEM creates a global address space that provides efficient and scalable communication for NVIDIA GPU clusters."},"popularity":{"tier":"top_1000"},"security":{"n_vulnerabilities":0},"version":"3.7.2"}