Share data center GPUs across virtual desktops, workstations and compute VMs.
NVIDIA virtual GPU (vGPU) software creates virtual GPUs from physical data center GPUs, so many virtual machines can share one GPU, or one VM can use several. Users get near bare-metal performance while IT keeps desktops, data and security centralised.
The latest vGPU release adds support for the RTX PRO 4500 Blackwell Server Edition, and integrates with the leading virtualization platforms.
Accelerated applications run in VMs with performance close to physical workstations.
Divide one GPU across many VMs or assign multiple GPUs to a single VM.
Offloading graphics from CPUs raises users per server and lowers TCO.
Regular releases with security features and professional support.
Integrated with Citrix, Nutanix and VMware platforms.
| NVIDIA RTX Virtual Workstation (vWS) | For designers, engineers, creators and AI developers |
|---|---|
| NVIDIA Virtual PC (vPC) and Virtual Applications (vApps) | For knowledge workers and published applications |
| Hypervisors | Citrix, Nutanix, VMware and other supported platforms |
|---|---|
| Supported GPUs | NVIDIA data center GPUs, including RTX PRO Blackwell Server Edition and L40S |
| Licensing | Enterprise licensing; free 90-day evaluation |
Production-grade software platform with NIM microservices, frameworks and enterprise support for AI.
Complete data center building blocks: compute, storage, networking, liquid cooling, power and management software.
Centralised IT management for ASUS servers — monitoring, firmware updates and remote control at scale.
Tell us your workload and timeline for the NVIDIA Virtual GPU (vGPU). We reply within one business day with a configuration, pricing and lead time.