Iris Coleman
Mar 20, 2025 03:44
NVIDIA’s vGPU 18.0 release enhances AI capabilities on virtual platforms, supporting Microsoft Windows Server 2025 and Proxmox VE, and introduces new AI toolkits for developers.
NVIDIA has unveiled its Virtual GPU (vGPU) 18.0 technology, marking a significant advancement in AI capabilities within Virtual Desktop Infrastructure (VDI). According to NVIDIA, this release aims to enhance productivity, security, and performance by leveraging AI-driven workloads across virtualized environments.
Enhanced Support for Microsoft Windows and Proxmox VE
One of the key features of the vGPU 18.0 update is its support for Microsoft Windows Server 2025. This integration allows users to utilize advanced management features such as GPU partitioning and live migration, facilitating efficient workload management and maintenance without disrupting virtual machines. Additionally, the update supports the Windows Subsystem for Linux (WSL), enabling seamless operation of Linux distributions alongside Windows.
Furthermore, the release includes support for Proxmox VE, an open-source server management platform. This addition to the NVIDIA vGPU supported hypervisors allows users to manage demanding Linux and Windows application workloads with software-defined GPU resources. Daire Byrne, global head of systems at DNEG, praised the seamless integration of NVIDIA vGPU with Proxmox VE, highlighting its ability to efficiently manage GPU utilization and support compute-intensive workloads.
New AI Toolkits and Industry Collaborations
NVIDIA has also introduced new AI Virtual Workstation (vWS) toolkits, aimed at accelerating AI development through VDI. The latest addition, the Fine-Tuning and Customizing LLMs toolkit, enables developers to use parameter-efficient fine-tuning techniques like low-rank adaptation (LoRA) to attach custom adapters to a single large language model (LLM) for various use cases.
The company has partnered with Citrix to deliver AI Virtual Workstations, allowing industries such as healthcare, financial services, and government to expand their AI-driven business objectives. This collaboration enhances the capability of Citrix-powered virtualized environments to run AI workloads while ensuring centralized data management and security.
Broader GPU Compatibility and Upcoming Enhancements
The vGPU 18.0 release also expands heterogeneous vGPU support, enabling a single GPU to handle multiple vGPU profiles with varying frame buffer sizes concurrently. This feature is now extended to NVIDIA Turing and Volta GPUs, optimizing GPU utilization across different generations of accelerated computing.
Looking ahead, NVIDIA announced the upcoming support for its RTX PRO 6000 Blackwell Server Edition GPU, expected in the latter half of 2025. This GPU, revealed at NVIDIA GTC 2025, promises to dynamically partition compute resources for diverse workloads, from graphics to AI, making it a versatile solution for data centers.
Overall, NVIDIA’s vGPU 18.0 release is set to transform virtualized environments by enhancing AI capabilities and supporting a broader range of platforms and applications. Businesses and developers are encouraged to explore these new features to maximize the potential of accelerated computing.
Image source: Shutterstock
