The NVIDIA Tesla V100 ultra-high-end GPU is the most advanced data center GPU ever built. Powered by NVIDIA Volta, the revolutionary Tesla V100 is ideal for accelerating the most demanding double-precision compute workflows and makes an ideal upgrade path from P100. Compress design cycles and accelerate time-to-market with up to 80 percent faster rendering performance1 versus Tesla P100.1 Enable engineers and designers to visualize their designs and perform complex simulations in real time, from any device, using the world’s most powerful virtual workstation.
Run NVIDIA Virtualization Software on NVIDIA GPUs
NVIDIA virtual GPU (vGPU) software runs on NVIDIA® GPUs and is based on NVIDIA Turing™, Volta™, Pascal™, and Maxwell™ GPU architectures. Match your needs with the right GPU below.
Performance Optimized
NVIDIA TESLA V100
1 Comparing a single virtual machine (VM) on NVIDIA Tesla V100-16Q with a single VM on NVIDIA Tesla P100-16Q using V-Ray 2015.
NVIDIA Quadro RTX 8000
The Quadro RTX 8000 comes with 48 GB memory – double the framebuffer of RTX 6000 enabling designers and artists to work with the largest and most complex ray tracing and visual computing workloads. When combined with NVIDIA Quadro vDWS software, the RTX 8000 enables the ultimate flexibility - powering virtual design workstations and render nodes to propel the creative workflow.
NVIDIA Quadro RTX 6000
With up to 75% graphics performance improvement versus Tesla P40, Quadro RTX 6000, when combined with NVIDIA Quadro vDWS software, enables artists to access high-powered virtual design workstations and render nodes to speed the design workflow and arrive at their best creations, faster. RT cores along with large frame buffer, and multiple profile sizes give artists and designers the flexibility to run the most demanding workloads from the data center. RTX 6000 enables organizations to virtualize both accelerated graphics and single-precision compute (NVIDIA CUDA® and OpenCL) workloads and harness the power of the industry’s most powerful rendering platform — NVIDIA RTX.
NVIDIA TESLA P40
The Tesla P40 mid-range to high-end graphics accelerator delivers up to 2X the professional graphics performance2 when compared to Tesla M60. Its large frame buffer, increased number of supported profiles, and larger profile sizes ensure you can support more users, including your most demanding ones. Tesla P40 enables organizations to virtualize both accelerated graphics and single-precision compute (NVIDIA CUDA® and OpenCL) workloads for every vGPU.
2 Comparing a single VM on NVIDIA Tesla M60-8Q vs. a single VM on NVIDIA Tesla P40-24Q and based on SPECviewperf 12.1 benchmark.
NVIDIA TESLA T4
With up to 2x the framebuffer versus P4 and 2x performance versus M60, the NVIDIA Tesla T4 is an ideal solution for entry to high-end 3D design and engineering workflows with NVIDIA Quadro vDWS software. With its single slot, low profile form factor and a mere 70 W power consumption, the Tesla T4 is a great option to achieve maximum GPU density per server node.
Density Optimized
NVIDIA TESLA M10
Tesla M10 is designed for data centers needing graphics acceleration for high-density virtual desktop environments to meet the needs of the modern digital workplace. It’s also ideal for organizations migrating to Windows 10—the most graphics-intensive operating system to date—or deploying additional infrastructure for disaster recovery and high availability due to business or regulatory requirements. It's a dual-slot PCI Express form factor for rack and tower servers, capable of supporting up to 64 concurrent users (512 MB profile).
Blade Optimized
NVIDIA TESLA P6
Tesla P6 is designed for blade servers and supports multiple data center workloads, including deep learning, high-performance computing, and graphics virtualization for demanding workloads. It delivers higher graphics performance, improved energy efficiency, and increased user density when compared to the Tesla M6, making it an ideal upgrade path. Tesla P6 comes in a mobile PCI Express module (MXM) form factor running at less than 90 W for high-density data centers with blade servers and converged infrastructure.
Additional GPUs Supported for Virtualization
COMPARE NVIDIA GPUs RECOMMENDED FOR VIRTUALIZATION
NVIDIA GPUs Recommended for Virtualization
| TESLA V100 | QUADRO RTX 8000 | QUADRO RTX 6000 | TESLA P40 | TESLA T4 | TESLA M10 | TESLA P6 | |
|---|---|---|---|---|---|---|---|
| GPU | 1 NVIDIA Volta GPU | 1 NVIDIA Turing GPU | 1 NVIDIA Turing GPU | 1 NVIDIA Pascal GPU | 1 NVIDIA Turing GPU | 4 NVIDIA Maxwell GPUs | 1 NVIDIA Pascal GPU |
| CUDA CORES | 5,120 | 4,608 | 4,608 | 3,840 | 2,560 | 2,560 (640 per GPU) | 2,048 |
| MEMORY SIZE | 16 GB HBM2 | 48 GB GDDR6 | 24 GB GDDR6 | 24 GB GDDR5 | 16 GB GDDR6 | 32 GB GDDR5 (8 GB per GPU) | 16 GB GDDR5 |
| USE CASES | High-end professional graphics users; includes use for double-precision compute workloads (3D models and design workflows, intensive CAE simulations) | Powerful RTX rendering platform for high-end professional graphics users (rendering and creative complex designs); includes single-precision compute workloads | Mid-range to high-end professional graphics users; includes single-precision compute workloads (rendering and creative complex designs); | Entry to mid-range professional graphics users including deep learning inference and rendering workloads, RTX-capable, 2 T4s are a suitable upgrade path from a single M60, or upgrade from a single P4 to a single T4 | Best density and performance for knowledge worker workloads | Ideal for blade server form factors and upgrades from M6 | |