Recommended NVIDIA GPUs for NVIDIA RTX vWS

Table 1 lists the hardware specifications for the latest-generation NVIDIA data center GPUs recommended for NVIDIA RTX Virtual Workstation (vWS). For a complete list of NVIDIA GPUs recommended for virtualization, see NVIDIA GPUs for virtualization.

These GPUs, based on the NVIDIA Blackwell, Ada Lovelace, and Ampere architectures, include second-, third-, and fourth-generation RT Cores, which accelerate ray tracing operations.

The GPUs in Table 1 are tested and supported with NVIDIA virtual GPU (vGPU) software. For the complete support matrices, see Virtual GPU Software Supported Products.

Table 1 NVIDIA GPUs Recommended for RTX vWS

	RTX PRO 6000 Blackwell Server Edition [1]	RTX PRO 4500 Blackwell Server Edition	L40S	L4	A40	A10
GPUs/ Board (Architecture)	Blackwell	Blackwell	Ada Lovelace	Ada Lovelace	Ampere	Ampere
Memory Size and Type	96GB GDDR7 with ECC	32GB GDDR7 with ECC	48GB GDDR6 with ECC	24GB GDDR6 with ECC	48GB GDDR6 with ECC	24GB GDDR6 with ECC
vGPU Profile Sizes (GB)	2GB, 3GB, 4GB, 6GB, 8GB, 12GB, 16GB, 24GB, 32GB, 48GB, 96GB	2GB, 3GB, 4GB, 8GB, 16GB, 32GB	1GB, 2GB, 3GB, 4GB, 6GB, 8GB, 12GB, 16GB, 24GB, 48GB	1GB, 2GB, 3GB, 4GB, 6GB, 8GB, 12GB, 24GB	1GB, 2GB, 3GB, 4GB, 6GB, 8GB, 12GB, 16GB, 24GB, 48GB	1GB, 2GB, 3GB, 4GB, 6GB, 8GB, 12GB, 24GB
MIG Support	Yes, up to 4 MIG slices	Yes, up to 2 MIG slices	No	No	No	No
NVLink Support	No	No	No	No	Yes	No
Form Factor	PCIe 5.0 Dual Slot FHFL	PCIe 5.0 Single Slot FHFL	PCIe 4.0 Dual Slot FHFL	PCIe 4.0 low-profile Single Slot	PCIe 4.0 Dual Slot FHFL	PCIe 4.0 Single Slot FHFL
Power (W)	600W	165W	350W	72W	300W	140W
Thermal	Passive	Passive	Passive	Passive	Passive	Passive
Optimized For [2]	Performance and Density	Density and Cost per User	Performance	Performance	Performance	Performance
Use Case	High-end 3D visualization applications, AI training and inference workloads, as well as vPC (VDI) deployments that benefit from high user density and excellent graphics performance	Data processing, graphics, video, and edge AI workloads. Well suited for vPC and entry-level vWS deployments	Accelerates deep learning and machine learning training and inference. Also supports high-end 3D design, creative workflows, and mixed workloads	End-to-end acceleration for next-gen AI and 3D graphics, rendering, and video	High-end 3D design and creative workflows with compute capabilities	Mainstream professional visualization on high-performance mid-range VWs

Note

Supported vGPU deployments require a certified server platform. Customers should refer to the NVIDIA Qualified Systems Catalog to verify that their host system is certified.

It is essential to resize your environment when switching to newer GPUs. For example, the NVIDIA T4 leverages ECC memory which is enabled by default. When enabled, ECC has a 1/15 overhead cost due to the need to use extra VRAM to store the ECC bits themselves; therefore, the amount of frame buffer usable by vGPU is reduced. Additional information for each hypervisor can be found in the respective NVIDIA documentation accessible here.

GPUs Supporting MIG-Backed vGPU

The NVIDIA RTX PRO 6000 Blackwell Server Edition and the NVIDIA RTX PRO 4500 Blackwell Server Edition support MIG-backed vGPU, enabling virtualized GPUs to be created from individual MIG slices and assigned to virtual machines. This model combines MIG’s hardware-level spatial partitioning with the temporal partitioning capabilities of vGPU, offering better QoS, and flexibility in how GPU resources are shared across workloads. The next section illustrates the benefits of MIG-backed vGPU with benchmark results for professional graphics workloads. More information on MIG-backed vGPU is available here.

NVIDIA RTX PRO 6000 Blackwell Server Edition

The NVIDIA RTX PRO 6000 Blackwell Server Edition is designed to deliver top-tier AI and graphics performance for enterprise data centers. It features 96 GB of high-speed GDDR7 ECC memory, 24,064 CUDA cores, 752 fifth-generation Tensor Cores, and 188 fourth-generation RT Cores. This combination makes it ideal for a wide range of workloads, including AI inference, simulation, high-quality rendering, and advanced computing tasks.

With Universal MIG, the RTX PRO 6000 Blackwell Server Edition becomes the first data center GPU capable of supporting both compute and graphics workloads within MIG instances. This enables flexible, mixed-use environments that combine AI/ML compute with professional visualization or VDI workloads while maintaining MIG’s strict resource isolation and predictable performance.

The RTX PRO 6000 Blackwell Server Edition is a powerful solution for both vWS and vPC, offering scalability and user density, supporting up to 48 concurrent vGPUs per GPU. Its exceptional performance, flexibility, and efficiency make it an ideal solution for organizations consolidating professional visualization, AI, and VDI workloads.

NVIDIA RTX PRO 4500 Blackwell Server Edition

The NVIDIA RTX PRO 4500 Blackwell Server Edition is designed to deliver efficient multi-workload acceleration for enterprise data centers, edge environments, and cloud deployments. It features 32 GB of high-speed GDDR7 memory, 10,496 CUDA cores, fifth-generation Tensor Cores, and fourth-generation RT Cores. This combination makes it well suited for a broad range of workloads, including virtualized data processing, graphics, video, and AI-accelerated workloads.

It also supports Universal MIG, enabling up to 2 isolated GPU instances for both compute and graphics workloads with predictable performance and strict resource guarantees.

The RTX PRO 4500 Blackwell Server Edition is a strong solution for vPC and entry-level vWS deployments, where user density, power efficiency, and cost optimization are important. Its 165 W single-slot design, support for up to two MIG-backed instances, and balanced performance across graphics and compute workloads make it a scalable and cost-effective option for organizations modernizing virtual desktop infrastructure.

NVIDIA L40S

The NVIDIA® L40S is the highest-performance Ada GPU for AI inference, AI training, and compute-intensive workloads, while also delivering excellent visual computing performance. Based on the NVIDIA Ada Lovelace GPU architecture, it provides exceptional performance for both advanced visual computing and AI workloads in data center and edge deployments. Featuring 142 third-generation RT Cores and 568 fourth-generation Tensor Cores with FP8 support, it accelerates real-time ray tracing, deep learning training and inference, generative AI workloads, and simulation workflows. With 48GB of graphics memory, the L40S delivers outstanding performance across compute-intensive tasks, batch and real-time rendering, virtual workstations, and cloud gaming. When combined with NVIDIA RTX™ Virtual Workstation (vWS) software, it enables powerful, secure virtual workstations that can be accessed from any device.

NVIDIA L4

The NVIDIA Ada Lovelace L4 Tensor Core GPU delivers universal acceleration and energy efficiency for video, AI, virtual workstations, and graphics applications in the enterprise, in the cloud, and at the edge. And with NVIDIA’s AI platform and full-stack approach, L4 is optimized for video and inference at scale for a broad range of AI applications to deliver the best in personalized experiences. As the most efficient NVIDIA accelerator for mainstream use, servers equipped with L4 power up to 120X higher AI video performance over CPU solutions and 2.5X more generative AI performance, as well as over 4X more graphics performance than the previous GPU generation. L4’s versatility and energy-efficient, single-slot, low-profile form factor makes it ideal for edge, cloud, and enterprise deployments.

NVIDIA A40

Built on the RTX platform, the NVIDIA A40 GPU is uniquely positioned to power high-end virtual workstations running professional visualization applications, accelerating the most demanding graphics workloads. The second-generation RT Cores of the NVIDIA A40 enable it to deliver massive speedups for workloads such as photorealistic rendering of movie content, architectural design evaluations, and virtual prototyping of product designs. The NVIDIA A40 features 48 GB of frame buffer, but with the NVIDIA® NVLink® GPU interconnect, it can support up to 96 GB of frame buffer to power virtual workstations that support very large animations, files, or models. Although the NVIDIA A40 has 48 GB of frame buffer, the context switching limit per GPU limits the maximum number of users supported to 32.

The NVIDIA A40 is also suitable for running VDI workloads and compute workloads on the same infrastructure. Resource utilization can be increased by using common virtualized GPU accelerated server resources to run virtual desktops and workstations while users are logged on, and compute workloads after the users have logged off. Learn more from the NVIDIA whitepaper about Using NVIDIA Virtual GPUs to Power Mixed Workloads.

NVIDIA A10

The NVIDIA A10 is designed to provide cost-effective graphics performance for accelerating and optimizing the performance of mixed workloads. When combined with NVIDIA RTX vWS software, it accelerates graphics and video processing with AI on mainstream enterprise servers. Its second-generation RT Cores make the NVIDIA A10 ideal for mainstream professional visualization applications running on high-performance mid-range virtual workstations.

vGPU Profile Selection Considerations

With the NVIDIA Blackwell, Ada, and Ampere architectures, expanding workload requirements can be accommodated given the evolution of GPU memory and available compute resources.

Smaller profiles (for example, 1,2 or 3 GB) can quickly become insufficient for demanding tasks such as 3D visualization, AI inference, or data-intensive simulations. Using larger profiles helps ensure that each vGPU instance has the required GPU memory and channels to deliver consistent performance across a variety of use cases.

The Blackwell architecture introduces hardware and software improvements that address channel limitations through support for MIG; larger profiles also have more GPU channels, improving the ability to run more GPU-intensive applications. For more details on GPU channel limitations, see Understanding GPU Channels.

Important points to consider:

• Lightweight Workload Demands: Although smaller profiles (1,2 or 3 GB) are not recommended for graphics-intensive or compute-heavy workloads, they remain suitable for lightweight visualization tasks such as simple 3D model viewing, or schematic editing. These profiles are also useful in environments that prioritize user density over high graphics performance.

• Modern Workload Demands: Applications such as 3D modeling, rendering, AI inference, and simulation require significant GPU memory. Smaller profiles are inadequate for these applications, leading to frequent memory overflows and degraded performance.

• Channel Limitations: GPUs have a limited number of channels. Smaller profiles quickly exhaust these channels, preventing efficient parallel processing and leading to application errors.

• Performance Optimization: Larger profiles provide the necessary memory bandwidth and capacity to handle complex workloads efficiently. With larger profiles, you can run fewer vGPUs simultaneously, but they receive more channels, making it less likely to encounter channel limitations and ensuring smooth and consistent performance.

• Scalability and Stability: Investing in larger profiles not only meets current demands but also provides a buffer for future workload increases, reducing the need for frequent upgrades.

The following examples illustrate typical workstation workloads and applications best suited to each profile size.

Table 2 Example Profile Recommendations by Use Case

Profile Size	Example Use Cases
1 GB	Basic 2D visualization
2 GB	Entry-level CAD model viewing, schematic editing
3 GB	Small CAD assemblies, basic 3D modeling, simple content creation or animation tasks
4 GB	Basic 3D design, light graphics work, office applications
6 GB	Medium 3D design, simple modeling, image processing
8 GB	3D modeling, AI-enhanced productivity apps, video editing
12 GB	Medium size CAD models, rendering
16 GB	Complex CAD models, high-end 3D modeling and animation, simulation and analysis, VR/AR
24 GB	Data analytics, game development, 3D visualization
32 GB	Large-scale simulations, high-end rendering, data visualization
48 GB	Multi-application workloads, rendering, AI workloads
96 GB	Very large simulations, advanced rendering, AI training

For detailed configurations and additional guidelines, please refer to the NVIDIA vGPU User Guide.

---

[1]

The NVIDIA RTX PRO 6000 Blackwell Server Edition also offers a liquid-cooled version, supported starting with vGPU 20.0.

[2]

Performance-optimized GPUs are designed to maximize raw performance for a specific class of virtualized workload. They are typically recommended for the following classes of virtualized workload:

• High-end virtual workstations running professional visualization applications.

• Compute-intensive workloads such as artificial intelligence, deep learning, or data science workloads.

Density-optimized GPUs are designed to maximize the number of VDI users supported in a server. They are typically recommended for knowledge worker virtual desktop infrastructure (VDI) to run office productivity applications, streaming video, and the Windows OS.