Virtual GPU Types Reference#
Virtual GPU Types for Supported GPUs#
NVIDIA vGPU is available as a licensed product on supported NVIDIA GPUs. For a list of recommended server platforms and supported GPUs, consult the release notes for supported hypervisors at NVIDIA Virtual GPU Software Documentation.
NVIDIA A40 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA A40#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A40-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A40-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A40-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A40-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A40-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A40-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A40-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A40-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A40-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A40-1Q |
1024 |
32[3] |
30 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA A40#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A40-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A40-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A40-1B |
1024 |
32 |
32 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA A40#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
A40-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
A40-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
A40-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
A40-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
A40-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
A40-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
A40-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
A40-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
A40-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
A40-1A |
1024 |
32[3] |
32 |
1280×1024 |
1 |
NVIDIA A16 Virtual GPU Types#
Physical GPUs per board: 4
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA A16#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A16-16Q |
16384 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A16-8Q |
8192 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A16-4Q |
4096 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A16-2Q |
2048 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A16-1Q |
1024 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA A16#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A16-3B |
3072 |
5 |
4 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A16-2B |
2048 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A16-1B |
1024 |
16 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA A16#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
A16-16A |
16384 |
1 |
1 |
1280×1024 |
1 |
A16-8A |
8192 |
2 |
2 |
1280×1024 |
1 |
A16-4A |
4096 |
4 |
4 |
1280×1024 |
1 |
A16-2A |
2048 |
8 |
8 |
1280×1024 |
1 |
A16-1A |
1024 |
16 |
16 |
1280×1024 |
1 |
NVIDIA A10 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA A10#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A10-24Q |
24576 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A10-12Q |
12288 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A10-8Q |
8192 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A10-6Q |
6144 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A10-4Q |
4096 |
6 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A10-3Q |
3072 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A10-2Q |
2048 |
12 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A10-1Q |
1024 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA A10#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A10-3B |
3072 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A10-2B |
2048 |
12 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A10-1B |
1024 |
24 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA A10#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
A10-24A |
24576 |
1 |
1 |
1280×1024 |
1 |
A10-12A |
12288 |
2 |
2 |
1280×1024 |
1 |
A10-8A |
8192 |
3 |
2 |
1280×1024 |
1 |
A10-6A |
6144 |
4 |
4 |
1280×1024 |
1 |
A10-4A |
4096 |
6 |
4 |
1280×1024 |
1 |
A10-3A |
3072 |
8 |
8 |
1280×1024 |
1 |
A10-2A |
2048 |
12 |
8 |
1280×1024 |
1 |
A10-1A |
1024 |
24 |
16 |
1280×1024 |
1 |
NVIDIA A2 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA A2#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A2-16Q |
16384 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A2-8Q |
8192 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
A2-4Q |
4096 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
A2-2Q |
2048 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
A2-1Q |
1024 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA A2#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
A2-3B |
3072 |
5 |
4 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A2-2B |
2048 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
A2-1B |
1024 |
16 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA A2#
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
A2-16A |
16384 |
1 |
1 |
1280×1024 |
1 |
A2-8A |
8192 |
2 |
2 |
1280×1024 |
1 |
A2-4A |
4096 |
4 |
4 |
1280×1024 |
1 |
A2-2A |
2048 |
8 |
8 |
1280×1024 |
1 |
A2-1A |
1024 |
16 |
16 |
1280×1024 |
1 |
NVIDIA L40 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA L40#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L40-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L40-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L40-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L40-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L40-1Q |
1024 |
32[4] |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA L40#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L40-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L40-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L40-1B |
1024 |
32 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA L40#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
L40-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
L40-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
L40-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
L40-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
L40-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
L40-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
L40-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
L40-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
L40-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
L40-1A |
1024 |
32[4] |
16 |
1280×1024 |
1 |
NVIDIA L40S Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA L40S#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L40S-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40S-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40S-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40S-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40S-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L40S-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L40S-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L40S-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L40S-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L40S-1Q |
1024 |
32[5] |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA L40S#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L40S-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L40S-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L40S-1B |
1024 |
32 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA L40S#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
L40S-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
L40S-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
L40S-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
L40S-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
L40S-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
L40S-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
L40S-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
L40S-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
L40S-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
L40S-1A |
1024 |
32[5] |
16 |
1280×1024 |
1 |
NVIDIA L20 and NVIDIA L20 Liquid Cooled Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
The virtual GPU types for the NVIDIA L20 and NVIDIA L20 liquid cooled GPUs are identical.
Q-Series Virtual GPU Types for NVIDIA L20 and NVIDIA L20 Liquid Cooled#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L20-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L20-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L20-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L20-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L20-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L20-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L20-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L20-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L20-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L20-1Q |
1024 |
32[6] |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA L20 and NVIDIA L20 Liquid Cooled#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L20-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L20-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L20-1B |
1024 |
32 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA L20 and NVIDIA L20 Liquid Cooled#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
L20-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
L20-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
L20-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
L20-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
L20-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
L20-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
L20-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
L20-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
L20-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
L20-1A |
1024 |
32[6] |
16 |
1280×1024 |
1 |
NVIDIA L4 Virtual GPU Types#
Physical GPUs per board: 1
Q-Series Virtual GPU Types for NVIDIA L4#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L4-24Q |
24576 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L4-12Q |
12288 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L4-8Q |
8192 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L4-6Q |
6144 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L4-4Q |
4096 |
6 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L4-3Q |
3072 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L4-2Q |
2048 |
12 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L4-1Q |
1024 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA L4#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L4-3B |
3072 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L4-2B |
2048 |
12 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L4-1B |
1024 |
24 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA L4#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
L4-24A |
24576 |
1 |
1 |
1280×1024 |
1 |
L4-12A |
12288 |
2 |
2 |
1280×1024 |
1 |
L4-8A |
8192 |
3 |
2 |
1280×1024 |
1 |
L4-6A |
6144 |
4 |
4 |
1280×1024 |
1 |
L4-4A |
4096 |
6 |
4 |
1280×1024 |
1 |
L4-3A |
3072 |
8 |
8 |
1280×1024 |
1 |
L4-2A |
2048 |
12 |
8 |
1280×1024 |
1 |
L4-1A |
1024 |
24 |
16 |
1280×1024 |
1 |
NVIDIA L2 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA L2#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L2-24Q |
24576 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L2-12Q |
12288 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L2-8Q |
8192 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
L2-6Q |
6144 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L2-4Q |
4096 |
6 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
L2-3Q |
3072 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L2-2Q |
2048 |
12 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
L2-1Q |
1024 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA L2#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
L2-3B |
3072 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L2-2B |
2048 |
12 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
L2-1B |
1024 |
24 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA L2#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
L2-24A |
24576 |
1 |
1 |
1280×1024 |
1 |
L2-12A |
12288 |
2 |
2 |
1280×1024 |
1 |
L2-8A |
8192 |
3 |
2 |
1280×1024 |
1 |
L2-6A |
6144 |
4 |
4 |
1280×1024 |
1 |
L2-4A |
4096 |
6 |
4 |
1280×1024 |
1 |
L2-3A |
3072 |
8 |
8 |
1280×1024 |
1 |
L2-2A |
2048 |
12 |
8 |
1280×1024 |
1 |
L2-1A |
1024 |
24 |
16 |
1280×1024 |
1 |
NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
The full virtual GPU type names include the prefix NVIDIA RTX PRO 6000 Blackwell, for example NVIDIA RTX PRO 6000 Blackwell DC-4-96Q.
This GPU supports MIG-backed virtual GPUs and time-sliced virtual GPUs.
MIG-Backed Q-Series Virtual GPU Types for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
DC-4-96Q MIG 4g.96gb+gfx |
96 |
1 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-48Q MIG 2g.48gb+gfx |
48 |
1 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-4-48Q MIG 4g.96gb+gfx |
48 |
2 |
4 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-4-32Q MIG 4g.96gb+gfx |
32 |
3 |
4 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-24Q MIG 1g.24gb+gfx |
24 |
1 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-24Q MIG 2g.48gb+gfx |
24 |
2 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-4-24Q MIG 4g.96gb+gfx |
24 |
4 |
4 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-16Q MIG 2g.48gb+gfx |
16 |
3 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-12Q MIG 1g.24gb+gfx |
12 |
2 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-12Q MIG 2g.48gb+gfx |
12 |
4 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-8Q MIG 1g.24gb+gfx |
8 |
3 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-6Q MIG 1g.24gb+gfx |
6 |
4 |
1 |
1 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
||||||
DC-1-4Q MIG 1g.24gb+gfx |
4 |
6 |
1 |
1 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
||||||
DC-1-3Q MIG 1g.24gb+gfx |
3 |
8 |
1 |
1 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
||||||
3840×2400 or lower |
4 |
||||||
DC-1-2Q MIG 1g.24gb+gfx |
2 |
12 |
1 |
1 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
||||||
3840×2400 or lower |
4 |
Time-Sliced Q-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-96Q |
96 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-48Q |
48 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-32Q |
32 |
3 |
3 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-24Q |
24 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-16Q |
16 |
6 |
6 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-12Q |
12 |
8 |
8 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-8Q |
8 |
12 |
12 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-6Q |
6 |
16 |
16 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
DC-4Q |
4 |
24 |
24 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
DC-3Q |
3 |
32 |
32 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
MIG-Backed B-Series Virtual GPU Types for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Max. Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
DC-1-3B MIG 1g.24gb+gfx |
3 |
8 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
||||||
DC-1-2B MIG 1g.24gb+gfx |
2 |
12 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
Time-Sliced B-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-3B |
3 |
32 |
32 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
DC-2B |
2 |
32 |
32 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
MIG-Backed A-Series Virtual GPU Types for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Max. Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-4-96A MIG 4g.96gb+gfx |
96 |
1 |
4 |
4 |
1280×1024 |
1 |
DC-2-48A MIG 2g.48gb+gfx |
48 |
1 |
2 |
2 |
1280×1024 |
1 |
DC-4-48A MIG 4g.96gb+gfx |
48 |
2 |
4 |
1 |
1280×1024 |
1 |
DC-4-32A MIG 4g.96gb+gfx |
32 |
3 |
4 |
1 |
1280×1024 |
1 |
DC-1-24A MIG 1g.24gb+gfx |
24 |
1 |
1 |
1 |
1280×1024 |
1 |
DC-2-24A MIG 2g.48gb+gfx |
24 |
2 |
2 |
1 |
1280×1024 |
1 |
DC-4-24A MIG 4g.96gb+gfx |
24 |
4 |
2 |
1 |
1280×1024 |
1 |
DC-2-16A MIG 2g.48gb+gfx |
16 |
3 |
2 |
1 |
1280×1024 |
1 |
DC-1-12A MIG 1g.24gb+gfx |
12 |
2 |
1 |
1 |
1280×1024 |
1 |
DC-2-12A MIG 2g.48gb+gfx |
12 |
4 |
2 |
1 |
1280×1024 |
1 |
DC-1-8A MIG 1g.24gb+gfx |
8 |
3 |
1 |
1 |
1280×1024 |
1 |
DC-1-6A MIG 1g.24gb+gfx |
6 |
4 |
1 |
1 |
1280×1024 |
1 |
DC-1-4A MIG 1g.24gb+gfx |
4 |
6 |
1 |
1 |
1280×1024 |
1 |
DC-1-3A MIG 1g.24gb+gfx |
3 |
8 |
1 |
1 |
1280×1024 |
1 |
DC-1-2A MIG 1g.24gb+gfx |
2 |
12 |
1 |
1 |
1280×1024 |
1 |
Time-Sliced A-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid Cooled#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
DC-96A |
96 |
1 |
1 |
1280×1024 |
1 |
DC-48A |
48 |
2 |
2 |
1280×1024 |
1 |
DC-32A |
32 |
3 |
3 |
1280×1024 |
1 |
DC-24A |
24 |
4 |
4 |
1280×1024 |
1 |
DC-16A |
16 |
6 |
6 |
1280×1024 |
1 |
DC-12A |
12 |
8 |
8 |
1280×1024 |
1 |
DC-8A |
8 |
12 |
12 |
1280×1024 |
1 |
DC-6A |
6 |
16 |
16 |
1280×1024 |
1 |
DC-4A |
4 |
24 |
24 |
1280×1024 |
1 |
DC-3A |
3 |
32 |
32 |
1280×1024 |
1 |
Since 20.2: NVIDIA RTX PRO 5000 Blackwell Workstation Edition 72 GB Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
The full virtual GPU type names include the prefix NVIDIA RTX PRO 5000 72GB, for example NVIDIA RTX PRO 5000 72GB Blackwell-12Q. The tables in this section use the short form of each name, for example Blackwell-12Q.
This GPU supports MIG-backed virtual GPUs and time-sliced virtual GPUs.
This GPU supports only Q-series virtual GPU types for NVIDIA RTX Virtual Workstation and B-series virtual GPU types for NVIDIA Virtual PC. Because this GPU is supported only in workstation systems, no A-series virtual GPU types are provided.
MIG-Backed Q-Series Virtual GPU Types for NVIDIA RTX PRO 5000 Blackwell Workstation Edition 72 GB#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
Blackwell-1-2Q MIG 1g.36gb+gfx |
2 |
18 |
1 |
1 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
||||||
3840×2400 or lower |
4 |
||||||
Blackwell-1-3Q MIG 1g.36gb+gfx |
3 |
12 |
1 |
1 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
||||||
3840×2400 or lower |
4 |
||||||
Blackwell-1-4Q MIG 1g.36gb+gfx |
4 |
9 |
1 |
1 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
||||||
Blackwell-1-6Q MIG 1g.36gb+gfx |
6 |
6 |
1 |
1 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
||||||
Blackwell-1-9Q MIG 1g.36gb+gfx |
9 |
4 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-1-12Q MIG 1g.36gb+gfx |
12 |
3 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-1-18Q MIG 1g.36gb+gfx |
18 |
2 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-1-36Q MIG 1g.36gb+gfx |
36 |
1 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-2-18Q MIG 2g.72gb+gfx |
18 |
4 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-2-24Q MIG 2g.72gb+gfx |
24 |
3 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-2-36Q MIG 2g.72gb+gfx |
36 |
2 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
Blackwell-2-72Q MIG 2g.72gb+gfx |
72 |
1 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
Time-Sliced Q-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 5000 Blackwell Workstation Edition 72 GB#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
Blackwell-72Q |
72 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-36Q |
36 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-24Q |
24 |
3 |
3 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-18Q |
18 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-12Q |
12 |
6 |
6 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-9Q |
9 |
8 |
8 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
Blackwell-6Q |
6 |
12 |
12 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
Blackwell-4Q |
4 |
18 |
18 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
Blackwell-3Q |
3 |
24 |
24 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
Blackwell-2Q |
2 |
32 |
32 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
MIG-Backed B-Series Virtual GPU Types for NVIDIA RTX PRO 5000 Blackwell Workstation Edition 72 GB#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Max. Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
Blackwell-1-3B MIG 1g.36gb+gfx |
3 |
12 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
||||||
Blackwell-1-2B MIG 1g.36gb+gfx |
2 |
18 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
Time-Sliced B-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 5000 Blackwell Workstation Edition 72 GB#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
Blackwell-3B |
3 |
24 |
24 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
Blackwell-2B |
2 |
32 |
32 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
NVIDIA RTX PRO 4500 Blackwell Server Edition Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
The full virtual GPU type names include the prefix NVIDIA RTX PRO 4500 Blackwell, for example NVIDIA RTX PRO 4500 Blackwell DC-2-32Q.
This GPU supports MIG-backed virtual GPUs and time-sliced virtual GPUs.
MIG-Backed Q-Series Virtual GPU Types for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
DC-2-32Q 2g.32gb+gfx |
32 |
1 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-16Q 1g.16gb+gfx |
16 |
1 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-16Q 2g.32gb+gfx |
16 |
2 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-8Q 1g.16gb+gfx |
8 |
2 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-2-8Q 2g.32gb+gfx |
8 |
4 |
2 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
||||||
DC-1-4Q 1g.16gb+gfx |
4 |
4 |
1 |
1 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
||||||
DC-1-2Q 1g.16gb+gfx |
2 |
8 |
1 |
1 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
||||||
3840×2400 or lower |
4 |
Time-Sliced Q-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-32Q |
32 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-16Q |
16 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-8Q |
8 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
DC-4Q |
4 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
MIG-Backed B-Series Virtual GPU Types for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Available Pixels |
Max. Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|---|
DC-1-3B MIG 1g.16gb+gfx |
3 |
5 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
||||||
DC-1-2B MIG 1g.16gb+gfx |
2 |
8 |
1 |
1 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
||||||
3840×2160 |
2 |
||||||
2560×1600 or lower |
4 |
Time-Sliced B-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-3B |
3 |
10 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
DC-2B |
2 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
MIG-Backed A-Series Virtual GPU Types for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
For details of GPU instance profiles, see NVIDIA Multi-Instance GPU User Guide.
Each virtual GPU type is supported only on its corresponding GPU instance profile.
Virtual GPU Type and Corresponding GPU Instance Profile |
Frame Buffer (GB) |
Max. vGPUs per GPU Instance |
Slices per vGPU |
Compute Instances per vGPU |
Max. Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
DC-2-32A 2g.32gb+gfx |
32 |
1 |
2 |
2 |
1280×1024 |
1 |
DC-1-16A 1g.16gb+gfx |
16 |
1 |
1 |
1 |
1280×1024 |
1 |
DC-2-16A 2g.32gb+gfx |
16 |
2 |
2 |
1 |
1280×1024 |
1 |
DC-1-8A 1g.16gb+gfx |
8 |
2 |
1 |
1 |
1280×1024 |
1 |
DC-2-8A 2g.32gb+gfx |
8 |
4 |
2 |
1 |
1280×1024 |
1 |
DC-1-4A 1g.16gb+gfx |
4 |
4 |
1 |
1 |
1280×1024 |
1 |
DC-1-2A 1g.16gb+gfx |
2 |
8 |
1 |
1 |
1280×1024 |
1 |
Time-Sliced A-Series Virtual GPU Types on Single-Instance GPUs for NVIDIA RTX PRO 4500 Blackwell Server Edition#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (GB) |
Max. vGPUs per GPU in Equal-Size Mode |
Max. vGPUs per GPU in Mixed-Size Mode |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
DC-32A |
32 |
1 |
1 |
1280×1024 |
1 |
DC-16A |
16 |
2 |
2 |
1280×1024 |
1 |
DC-8A |
8 |
4 |
4 |
1280×1024 |
1 |
DC-4A |
4 |
8 |
8 |
1280×1024 |
1 |
NVIDIA RTX 6000 Ada Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX 6000 Ada#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 6000 Ada-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTX 6000 Ada-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTX 6000 Ada-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTX 6000 Ada-1Q |
1024 |
32[7] |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX 6000 Ada#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 6000 Ada-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 6000 Ada-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 6000 Ada-1B |
1024 |
32 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX 6000 Ada#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTX 6000 Ada-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
RTX 6000 Ada-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
RTX 6000 Ada-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
RTX 6000 Ada-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
RTX 6000 Ada-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
RTX 6000 Ada-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
RTX 6000 Ada-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
RTX 6000 Ada-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
RTX 6000 Ada-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
RTX 6000 Ada-1A |
1024 |
32[7] |
16 |
1280×1024 |
1 |
NVIDIA RTX 5880 Ada Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX 5880 Ada#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 5880 Ada-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTX 5880 Ada-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTX 5880 Ada-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTX 5880 Ada-1Q |
1024 |
32[8] |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX 5880 Ada#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 5880 Ada-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 5880 Ada-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 5880 Ada-1B |
1024 |
32 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX 5880 Ada#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTX 5880 Ada-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
RTX 5880 Ada-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
RTX 5880 Ada-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
RTX 5880 Ada-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
RTX 5880 Ada-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
RTX 5880 Ada-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
RTX 5880 Ada-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
RTX 5880 Ada-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
RTX 5880 Ada-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
RTX 5880 Ada-1A |
1024 |
32[8] |
16 |
1280×1024 |
1 |
NVIDIA RTX 5000 Ada Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX 5000 Ada#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 5000 Ada-32Q |
32768 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5000 Ada-16Q |
16384 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5000 Ada-8Q |
8192 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTX 5000 Ada-4Q |
4096 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTX 5000 Ada-2Q |
2048 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTX 5000 Ada-1Q |
1024 |
32 |
32 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX 5000 Ada#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTX 5000 Ada-3B |
3072 |
10 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 5000 Ada-2B |
2048 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTX 5000 Ada-1B |
1024 |
32 |
32 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX 5000 Ada#
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTX 5000 Ada-32A |
32768 |
1 |
1 |
1280×1024 |
1 |
RTX 5000 Ada-16A |
16384 |
2 |
2 |
1280×1024 |
1 |
RTX 5000 Ada-8A |
8192 |
4 |
4 |
1280×1024 |
1 |
RTX 5000 Ada-4A |
4096 |
8 |
8 |
1280×1024 |
1 |
RTX 5000 Ada-2A |
2048 |
16 |
16 |
1280×1024 |
1 |
RTX 5000 Ada-1A |
1024 |
32 |
32 |
1280×1024 |
1 |
NVIDIA RTX A6000 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX A6000#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA6000-48Q |
49152 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA6000-24Q |
24576 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA6000-16Q |
16384 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA6000-12Q |
12288 |
4 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA6000-8Q |
8192 |
6 |
4 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA6000-6Q |
6144 |
8 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA6000-4Q |
4096 |
12 |
8 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA6000-3Q |
3072 |
16 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA6000-2Q |
2048 |
24 |
16 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA6000-1Q |
1024 |
32[9] |
30 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX A6000#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA6000-3B |
3072 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA6000-2B |
2048 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA6000-1B |
1024 |
32 |
30 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX A6000#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTXA6000-48A |
49152 |
1 |
1 |
1280×1024 |
1 |
RTXA6000-24A |
24576 |
2 |
2 |
1280×1024 |
1 |
RTXA6000-16A |
16384 |
3 |
2 |
1280×1024 |
1 |
RTXA6000-12A |
12288 |
4 |
4 |
1280×1024 |
1 |
RTXA6000-8A |
8192 |
6 |
4 |
1280×1024 |
1 |
RTXA6000-6A |
6144 |
8 |
8 |
1280×1024 |
1 |
RTXA6000-4A |
4096 |
12 |
8 |
1280×1024 |
1 |
RTXA6000-3A |
3072 |
16 |
16 |
1280×1024 |
1 |
RTXA6000-2A |
2048 |
24 |
16 |
1280×1024 |
1 |
RTXA6000-1A |
1024 |
32[9] |
30 |
1280×1024 |
1 |
NVIDIA RTX A5500 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX A5500#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA5500-24Q |
24576 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5500-12Q |
12288 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5500-8Q |
8192 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5500-6Q |
6144 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA5500-4Q |
4096 |
6 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA5500-3Q |
3072 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA5500-2Q |
2048 |
12 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA5500-1Q |
1024 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX A5500#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA5500-3B |
3072 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA5500-2B |
2048 |
12 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA5500-1B |
1024 |
24 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX A5500#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTXA5500-24A |
24576 |
1 |
1 |
1280×1024 |
1 |
RTXA5500-12A |
12288 |
2 |
2 |
1280×1024 |
1 |
RTXA5500-8A |
8192 |
3 |
2 |
1280×1024 |
1 |
RTXA5500-6A |
6144 |
4 |
4 |
1280×1024 |
1 |
RTXA5500-4A |
4096 |
6 |
4 |
1280×1024 |
1 |
RTXA5500-3A |
3072 |
8 |
8 |
1280×1024 |
1 |
RTXA5500-2A |
2048 |
12 |
8 |
1280×1024 |
1 |
RTXA5500-1A |
1024 |
24 |
16 |
1280×1024 |
1 |
NVIDIA RTX A5000 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for NVIDIA RTX A5000#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA5000-24Q |
24576 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5000-12Q |
12288 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5000-8Q |
8192 |
3 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
RTXA5000-6Q |
6144 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA5000-4Q |
4096 |
6 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
RTXA5000-3Q |
3072 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA5000-2Q |
2048 |
12 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
RTXA5000-1Q |
1024 |
24 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for NVIDIA RTX A5000#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
RTXA5000-3B |
3072 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA5000-2B |
2048 |
12 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
RTXA5000-1B |
1024 |
24 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for NVIDIA RTX A5000#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
RTXA5000-24A |
24576 |
1 |
1 |
1280×1024 |
1 |
RTXA5000-12A |
12288 |
2 |
2 |
1280×1024 |
1 |
RTXA5000-8A |
8192 |
3 |
2 |
1280×1024 |
1 |
RTXA5000-6A |
6144 |
4 |
4 |
1280×1024 |
1 |
RTXA5000-4A |
4096 |
6 |
4 |
1280×1024 |
1 |
RTXA5000-3A |
3072 |
8 |
8 |
1280×1024 |
1 |
RTXA5000-2A |
2048 |
12 |
8 |
1280×1024 |
1 |
RTXA5000-1A |
1024 |
24 |
16 |
1280×1024 |
1 |
Tesla T4 Virtual GPU Types#
Physical GPUs per board: 1
The maximum number of vGPUs per board is the product of the maximum number of vGPUs per GPU and the number of physical GPUs per board.
Q-Series Virtual GPU Types for Tesla T4#
Intended use case: Virtual Workstations
Required license edition: vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
T4-16Q |
16384 |
1 |
1 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
T4-8Q |
8192 |
2 |
2 |
66355200 |
7680×4320 |
2 |
5120×2880 or lower |
4 |
|||||
T4-4Q |
4096 |
4 |
4 |
58982400 |
7680×4320 |
1 |
5120×2880 or lower |
4 |
|||||
T4-2Q |
2048 |
8 |
8 |
36864000 |
7680×4320 |
1 |
5120×2880 |
2 |
|||||
3840×2400 or lower |
4 |
|||||
T4-1Q |
1024 |
16 |
16 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
B-Series Virtual GPU Types for Tesla T4#
Intended use case: Virtual Desktops
Required license edition: vPC or vWS
These vGPU types support a maximum combined resolution based on the number of available pixels, which is determined by their frame buffer size. You can choose between using a small number of high resolution displays or a larger number of lower resolution displays with these vGPU types. The maximum number of displays per vGPU is based on a configuration in which all displays have the same resolution. For examples of configurations with a mixture of display resolutions, see Mixed Display Configurations for B-Series and Q-Series vGPUs.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Available Pixels |
Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|---|
T4-2B |
2048 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
T4-2B4[1] |
2048 |
8 |
8 |
18432000 |
5120×2880 |
1 |
3840×2400 |
2 |
|||||
3840×2160 |
2 |
|||||
2560×1600 or lower |
4 |
|||||
T4-1B |
1024 |
16 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
|||||
T4-1B4[1] |
1024 |
16 |
16 |
16384000 |
5120×2880 |
1 |
3840×2400 |
1 |
|||||
3840×2160 |
1 |
|||||
2560×1600 or lower |
4[2] |
A-Series Virtual GPU Types for Tesla T4#
Intended use case: Virtual Applications
Required license edition: vApps
These vGPU types support a single display with a fixed maximum resolution.
Virtual GPU Type |
Frame Buffer (MB) |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Maximum Display Resolution |
Virtual Displays per vGPU |
|---|---|---|---|---|---|
T4-16A |
16384 |
1 |
1 |
1280×1024 |
1 |
T4-8A |
8192 |
2 |
2 |
1280×1024 |
1 |
T4-4A |
4096 |
4 |
4 |
1280×1024 |
1 |
T4-2A |
2048 |
8 |
8 |
1280×1024 |
1 |
T4-1A |
1024 |
16 |
16 |
1280×1024 |
1 |
Mixed Display Configurations for B-Series and Q-Series vGPUs#
Mixed Display Configurations for B-Series vGPUs#
Virtual GPU Type |
Available Pixels |
Available Pixel Basis |
Maximum Displays |
Sample Mixed Display Configurations |
|---|---|---|---|---|
-3B |
18432000 |
2 3840×2400 displays |
4 |
1 3840×2400 display plus 2 2560×1600 displays |
-2B |
18432000 |
2 3840×2400 displays |
4 |
1 3840×2400 display plus 2 2560×1600 displays |
-2B4 |
18432000 |
2 3840×2400 displays |
4 |
1 3840×2400 display plus 2 2560×1600 displays |
-1B |
16384000 |
4 2560×1600 displays |
4 |
1 4096×2160 display plus 1 2560×1600 display |
-1B4 |
16384000 |
4 2560×1600 displays |
4 |
1 4096×2160 display plus 1 2560×1600 display |
Mixed Display Configurations for Q-Series vGPUs#
Virtual GPU Type |
Available Pixels |
Available Pixel Basis |
Maximum Displays |
Sample Mixed Display Configurations |
|---|---|---|---|---|
-8Q and above |
66355200 |
2 7680×4320 displays |
4 |
1 7680×4320 display plus 2 5120×2880 displays |
1 7680×4320 display plus 3 4096×2160 displays |
||||
-6Q |
58982400 |
4 5120×2880 displays |
4 |
1 7680×4320 display plus 1 5120×2880 display |
-4Q |
58982400 |
4 5120×2880 displays |
4 |
1 7680×4320 display plus 1 5120×2880 display |
-3Q |
35389440 |
4 4096×2160 displays |
4 |
1 5120×2880 display plus 2 4096×2160 displays |
-2Q |
35389440 |
4 4096×2160 displays |
4 |
1 5120×2880 display plus 2 4096×2160 displays |
-1Q |
17694720 |
2 4096×2160 displays |
4 |
1 4096×2160 display plus 2 2560×1600 displays |
vGPU Placements for GPUs#
The vGPU placements that a GPU supports depend on the total amount of frame buffer that the GPU has and whether the GPU is in mixed-size mode or equal-size mode.
In equal-size mode, the vGPU placement IDs are integer multiples of the placement size from zero to 1 fewer than the maximum vGPUs per GPU. For example, on a GPU with 40 GB of frame buffer, the maximum number of vGPUs with 10240 MB of frame buffer, which have a placement size of 10, is 4. The placement IDs for these vGPUs are as follows:
0 × 10 = 0
1 × 10 =10
2 × 10 = 20
3 × 10 = 30
vGPU Placements for GPUs with 96 GB of Frame Buffer#
The placement region size depends on the number GPU instances on the GPU.
vGPU placement IDs for GPUs with 96 GB of frame buffer are the same in mixed-size mode and in equal-size mode.
Note
For MIG-backed vGPUs, the placement IDs and placement region are local to the GPU instance, not the physical GPU. Therefore, multiple vGPUs on different GPU instances on the same GPU can have the same placement ID.
vGPU Placements on a Single-Instance GPU#
Placement region size: 96
A single-instance GPU is a GPU that is not partitioned into multiple GPU instances.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs |
|---|---|---|---|---|
98304 |
96 |
1 |
1 |
0 |
49152 |
48 |
2 |
2 |
0, 48 |
32768 |
32 |
3 |
3 |
0, 32, 64 |
24576 |
24 |
4 |
4 |
0, 24, 48, 72 |
16384 |
16 |
6 |
6 |
0, 16, 32, 48, 64, 80 |
12288 |
12 |
8 |
8 |
0, 12, 24, 36, 48, 60, 72, 84 |
8192 |
8 |
12 |
12 |
0, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88 |
6144 |
6 |
16 |
16 |
0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66, 72, 78, 84, 90 |
4096 |
4 |
24 |
24 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 64, 68, 72, 76, 80, 84, 88, 92 |
3072 |
3 |
32 |
32 |
0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51, 54, 57, 60, 63, 66, 69, 72, 75, 78, 81, 84, 87, 90, 93 |
2048 |
2 |
32 |
32 |
0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51, 54, 57, 60, 63, 66, 69, 72, 75, 78, 81, 84, 87, 90, 93 |
The following diagram shows the supported placements for each size of vGPU on a single-instance GPU with a total of 96 GB of frame buffer in mixed-size mode and in equal-size mode.

vGPU Placements on a Full-Partition GPU Instance (4g.96gb+gfx)#
Placement region size: 96
A full partition GPU instance is assigned all of the GPU’s frame buffer.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs |
|---|---|---|---|---|
98304 |
96 |
1 |
1 |
0 |
49152 |
48 |
2 |
2 |
0, 48 |
32768 |
32 |
3 |
3 |
0, 32, 64 |
24576 |
24 |
4 |
4 |
0, 24, 48, 72 |
The following diagram shows the supported placements for each size of vGPU on a full-partition GPU instance on a GPU with a total of 96 GB of frame buffer in mixed-size mode and in equal-size mode.

vGPU Placements on a Half-Partition GPU Instance (2g.48gb+gfx)#
Placement region size: 48
A half- partition GPU instance is assigned half of the GPU’s frame buffer.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs |
|---|---|---|---|---|
49152 |
48 |
1 |
1 |
0 |
24576 |
24 |
2 |
2 |
0, 24 |
16384 |
26 |
3 |
3 |
0, 16, 32 |
12288 |
12 |
4 |
4 |
0, 12, 24, 36 |
The following diagram shows the supported placements for each size of vGPU on a half-partition GPU instance on a GPU with a total of 96 GB of frame buffer in mixed-size mode and in equal-size mode.

vGPU Placements on a Quarter-Partition GPU Instance (1g.24gb+gfx)#
Placement region size: 48
A quarter- partition GPU instance is assigned one quarter of the GPU’s frame buffer.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs |
|---|---|---|---|---|
24576 |
24 |
1 |
1 |
0 |
12288 |
12 |
2 |
2 |
0, 12 |
8192 |
8 |
3 |
3 |
0, 8, 16 |
6144 |
6 |
4 |
4 |
0, 6, 12, 18 |
4096 |
4 |
6 |
6 |
0, 4, 8, 12, 16, 20 |
3072 |
3 |
8 |
8 |
0, 3, 6, 9, 12, 15, 18, 21 |
2048 |
2 |
12 |
12 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22 |
The following diagram shows the supported placements for each size of vGPU on a quarter-partition GPU instance on a GPU with a total of 96 GB of frame buffer in mixed-size mode and in equal-size mode.

vGPU Placements for GPUs with 94 GB of Frame Buffer#
Placement region size: 94
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
96256 |
94 |
1 |
0 |
1 |
0 |
48128 |
47 |
2 |
0, 47 |
2 |
0, 47 |
23552 |
23 |
4 |
0, 23, 46, 79 |
4 |
0, 24, 47, 71 |
15360 |
15 |
6 |
0, 15, 30, 45, 60, 75 |
4 |
0, 32, 47, 79 |
11264 |
11 |
8 |
0, 11, 22, 33, 44, 55, 66, 77 |
8 |
0, 12, 23, 36, 47, 59, 70, 83 |
6,144 |
6 |
15 |
0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66, 72, 78, 84 |
8 |
0, 17, 23, 41, 47, 64, 70, 88 |
4,096 |
4 |
23 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 64, 68, 72, 76, 80, 84, 88 |
16 |
0, 7, 11, 19, 23, 31, 35, 43, 47, 54, 58, 66, 70, 78, 82, 90 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 94 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 94 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 80 GB of Frame Buffer#
Placement region size: 80
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
81920 |
80 |
1 |
0 |
1 |
0 |
40960 |
40 |
2 |
0, 40 |
2 |
0, 40 |
20480 |
20 |
4 |
0, 20, 40, 60 |
4 |
0, 20, 40, 60 |
16384 |
16 |
5 |
0, 16, 32, 48, 64 |
4 |
0, 24, 40, 64 |
10240 |
10 |
8 |
0, 10, 20, 30, 40, 50, 60, 70 |
8 |
0, 10, 20, 30, 40, 50, 60, 70 |
8192 |
8 |
10 |
0, 8, 16, 24, 32, 40, 48, 56, 64, 72 |
8 |
0, 12, 20, 32, 40, 52, 60, 72 |
5120 |
5 |
16 |
0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 |
16 |
0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 |
4096 |
4 |
20 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 66, 70, 76 |
16 |
0, 6, 10, 16, 20, 26, 30, 36, 40, 46, 50, 56, 60, 66, 70, 76 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 80 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 80 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 64 GB of Frame Buffer#
Placement region size: 64
vGPU placement IDs for GPUs with 64 GB of frame buffer are the same in mixed-size mode and in equal-size mode.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs |
|---|---|---|---|---|
65536 |
64 |
1 |
1 |
0 |
32768 |
32 |
2 |
2 |
0, 32 |
16384 |
16 |
4 |
4 |
0, 16, 32, 48 |
8192 |
8 |
8 |
8 |
0, 8, 16, 24, 32, 40, 48, 56 |
4096 |
4 |
16 |
16 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 46, 60 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 64 GB of frame buffer in mixed-size mode and in equal-size mode.

vGPU Placements for GPUs with 48 GB of Frame Buffer#
Placement region size: 48
Note
When in mixed-size mode, the maximum number of vGPUs with 1024 MB of frame buffer allowed on GPUs based on the Ada Lovelace GPU architecture is lower than for other GPU architectures. As a result, the supported placement IDs for these vGPUs on GPUs based on the Ada Lovelace GPU architecture are different than for other GPU architectures.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
49152 |
48 |
1 |
0 |
1 |
0 |
24576 |
24 |
2 |
0, 24 |
2 |
0, 24 |
16384 |
16 |
3 |
0, 16, 32 |
2 |
0, 32 |
12288 |
12 |
4 |
0, 12, 24, 36 |
4 |
0, 12, 24, 36 |
8192 |
8 |
6 |
0, 8, 16, 24, 32, 40 |
4 |
0, 16, 24, 40 |
6144 |
6 |
8 |
0, 6, 12, 18, 24, 30, 36, 42 |
8 |
0, 6, 12, 18, 24, 30, 36, 42 |
4096 |
4 |
12 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44 |
8 |
0, 8, 12, 20, 24, 32, 36, 44 |
3072 |
3 |
16 |
0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45 |
16 |
0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45 |
2048 |
2 |
24 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46 |
16 |
0, 4, 6, 10, 12, 16, 18, 22, 24, 28, 30, 34, 36, 40, 42, 46 |
1024 |
1 |
32 |
0 through 31 |
GPU architectures except Ada Lovelace: 30 |
GPU architectures except Ada Lovelace: 0, 2, 3, 5, 6, 9, 11, 12, 14, 15, 17, 18, 20, 21, 23, 24, 26, 27, 29, 30, 33, 35, 36, 38, 39, 41, 42, 44, 45, 47 |
Ada Lovelace GPU architecture: 16 |
Ada Lovelace GPU architecture: 0, 5, 6, 11, 12, 17, 18, 23, 24, 29, 30, 35, 36, 41, 42, 47 |
The following diagram shows the supported placements for each size of vGPU on a GPU based on a GPU architecture except Ada Lovelace with a total of 48 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU based on the Ada Lovelace GPU architecture with a total of 48 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 48 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 40 GB of Frame Buffer#
Placement region size: 40
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
40960 |
40 |
1 |
0 |
1 |
0 |
20480 |
20 |
2 |
0, 20 |
2 |
0, 20 |
10240 |
10 |
4 |
0, 10, 20, 30 |
4 |
0, 10, 20, 30 |
8192 |
8 |
5 |
0, 8, 16, 24, 32 |
4 |
0, 12, 20, 32 |
5120 |
5 |
8 |
0, 5, 10, 15, 20, 25, 30, 35 |
8 |
0, 5, 10, 15, 20, 25, 30, 35 |
4096 |
4 |
10 |
0, 4, 8, 12, 16, 20, 24, 28, 32, 36 |
8 |
0, 6, 10, 16, 20, 26, 30, 36 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 40 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 40 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 32 GB of Frame Buffer#
Placement region size: 32
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
32768 |
32 |
1 |
0 |
1 |
0 |
16384 |
16 |
2 |
0, 16 |
2 |
0, 16 |
8192 |
8 |
4 |
0, 8, 16, 24 |
4 |
0, 8, 16, 24 |
4096 |
4 |
8 |
0, 4, 8, 12, 16, 20, 24, 28 |
8 |
0, 4, 8, 12, 16, 20, 24, 28 |
3072 |
3 |
10 |
0, 3, 6, 9, 12, 15, 18, 21, 24, 27 |
8 |
0, 5, 8, 13, 16, 21, 24, 29 |
2048 |
2 |
16 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30 |
16 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30 |
1024 |
1 |
32 |
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 |
32 |
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 32 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 32 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 24 GB of Frame Buffer#
Placement region size: 24
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
24576 |
24 |
1 |
0 |
1 |
0 |
12288 |
12 |
2 |
0, 12 |
2 |
0, 12 |
8192 |
8 |
3 |
0, 8, 16 |
2 |
0, 16 |
6144 |
6 |
4 |
0, 6, 12, 18 |
4 |
0, 6, 12, 18 |
4096 |
4 |
6 |
0, 4, 8, 12, 16, 20 |
4 |
0, 8, 12, 20 |
3072 |
3 |
8 |
0, 3, 6, 9, 12, 15, 18, 21 |
8 |
0, 3, 6, 9, 12, 15, 18, 21 |
2048 |
2 |
12 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22 |
8 |
0, 4, 6, 10, 12, 16, 18, 22 |
1024 |
1 |
24 |
0 through 23 |
16 |
0, 2, 3, 5, 6, 8, 9, 11, 12, 14, 15, 17, 18, 20, 21, 23 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 24 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 24 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 20 GB of Frame Buffer#
Placement region size: 20
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
20480 |
20 |
1 |
0 |
1 |
0 |
10240 |
10 |
2 |
0, 10 |
2 |
0, 10 |
5120 |
5 |
4 |
0, 5, 10, 15 |
4 |
0, 5, 10, 15 |
4096 |
4 |
5 |
0, 4, 8, 12, 16 |
4 |
0, 6, 10, 16 |
2048 |
2 |
10 |
0, 2, 4, 6, 8, 10, 12, 14, 16, 18 |
8 |
0, 3, 5, 8, 10, 13, 15, 18 |
1024 |
1 |
20 |
0 through 19 |
16 |
0, 1, 2, 4, 5, 6, 7, 9, 10, 11, 12, 14, 15, 16, 17, 19 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 20 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 20 GB of frame buffer in equal-size mode.

vGPU Placements for GPUs with 16 GB of Frame Buffer#
Placement region size: 16
vGPU placement IDs for GPUs with 16 GB of frame buffer are the same in mixed-size mode and in equal-size mode.
vGPU Size (MB of Frame Buffer) |
Placement Size |
Maximum vGPUs per GPU in Equal-Size Mode |
Supported Placement IDs in Equal-Size Mode |
Maximum vGPUs per GPU in Mixed-Size Mode |
Supported Placement IDs in Mixed-Size Mode |
|---|---|---|---|---|---|
16384 |
16 |
1 |
0 |
1 |
0 |
8192 |
8 |
2 |
0, 8 |
2 |
0, 8 |
4096 |
4 |
4 |
0, 4, 8, 12 |
4 |
0, 4, 8, 12 |
3072 |
3 |
5 |
0, 3, 6, 9, 12 |
4 |
0, 5, 8, 13 |
2048 |
2 |
8 |
0, 2, 4, 6, 8, 10, 12, 14 |
8 |
0, 2, 4, 6, 8, 10, 12, 14 |
1024 |
1 |
16 |
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
16 |
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 16 GB of frame buffer in mixed-size mode.

The following diagram shows the supported placements for each size of vGPU on a GPU with a total of 16 GB of frame buffer in equal-size mode.
