Horizon 7 Sizing and Consumption Model : 7.3 Infrastructure Blueprints : 7.3.1 vGPU
   
7.3.1 vGPU
One of the advantages of Horizon 7 is the ability to offer GPU (Graphical Processing Unit) accelerated workloads for 3D graphical capabilities or machine learning applications which is gaining in popularity.
Infrastructure blueprints may include NVIDIA GPU-enabled clusters using either the GRID 1.0 (K1 or K2) or GRID 2.0 (M60) GPU cards. These can be utilized by tenants with virtual desktops, or using RDSH sessions.
Table 9. NVIDIA GRID vGPU Profiles
Card
Physical GPUs per board
GPU Profiles
Frame Buffer
Max Displays per User
Max vGPUs per pGPU
Max vGPUs per board
NVIDIA GRID K2
2
K280Q
K260Q
K240Q
K220Q
4GB
2GB
1GB
512MB
4
4
2
2
1
2
4
8
2
4
8
16
NVIDIA GRID K1
1
K180Q
K160Q
K140Q
K120Q
4GB
2GB
1GB
512MB
4
4
2
2
1
2
4
8
4
8
16
32
NVIDIA
TESLA M60
GRID Virtual Workstation
2
M60-8Q
M60-4Q
M60-2Q
M60-1Q
M60-0Q
8GB
4GB
2GB
1GB
512MB
4
4
4
2
2
1
2
4
8
16
2
4
8
16
32
NVIDIA
TESLA M60 Virtual PC
2
M60-1B
M60-0B
1GB
512MB
4
2
8
16
16
32
NVIDIA
TESLA M60
Virtual Applications
2
M60-8A
M60-4A
M60-2A
M60-1A
8GB
4GB
2GB
1GB
1
1
1
1
1
2
4
8
2
4
8
16
NVIDIA TESLA M10
GRID Virtual Workstation
4
M10-8Q
M10-4Q
M10-2Q
M10-1Q
M10-0Q
8GB
4GB
2GB
1GB
512MB
4
4
4
2
2
1
2
4
8
16
4
8
16
32
64
NVIDIA
TESLA M10 Virtual PC
4
M10-1B
M10-0B
1GB
512MB
4
2
8
16
32
64
NVIDIA
TESLA M10
Virtual Applications
4
M10-8A
M10-4A
M10-2A
M10-1A
8GB
4GB
2GB
1GB
1
1
1
1
1
2
4
8
4
8
16
32