Product Range

Vendors

NVIDIA 289 3M 85 Aetina 21 Alta Labs 12 Atdec 402 Aver 19 BenQ 89 Brenthaven 1 Cambium Networks 3 Cherry 4 Chiq 13 Cisco 70 Connect Tech 27 D-Link 219 DELL 2 Everki 58 EVGA 1 FSP 4 Fujitsu 8 Gilkon 1 Gumdrop 55 Havis 26 Hikvision 60 HP 364 IKey 9 Infocase 124 ION 37 J5Create 42 LapCabby 22 Lenovo 117 Lexar 22 LIND 13 Logitech 27 Matrox 63 Microsoft 1 Miscellaneous Vendors 15 MSI 7 NetComm 7 OKI 215 Panasonic 299 Patchbox 62 Pelican 121 Philips 60 PICO 6 PowerShield 85 Rammount 8 Seagate 15 Sharp 90 Shintaro 102 Smartsign 1 Sony 63 SOTI 1 Supermicro 10 Synology 181 Team Group 33 Touchboard Mounting Systems 16 Ubiquiti 30 Unitek 43 Vertiv 239 Western Digital 11 Yamaha 10

Cloud Managed

No 3 Yes 18

Compute APIs

CUDA 11.6 DirectCompute OpenCL 3.0 4 CUDA 11.6 OpenCL 3.0 DirectCompute 2 CUDA 12.2 OpenCL 3.0 DirectCompute 2 CUDA 12.8 OpenCL 3.0 2 CUDA 12.8 OpenCL 3.0 DirectCompute 6

Cooling

Active 20 Passive 1

CUDA Cores

Display connectors

4 x DisplayPort 1.4a 4 4x DisplayPort 2.1 1 4x DisplayPort 2.1b 6 4x Mini DisplayPort 1.4a 5 4x mini DisplayPort 2.1b 2

Encode/decode engines

1x encode 1x decode (+AV1 decode) 1 1x encode 1x decode (+AV1 encode and decode) 1 1x encode 2x decode (+AV1 decode) 1 2x encode 2x decode (+AV1 encode and decode) 4 3x encode 3x decode (+AV1 encode and decode) 1 3x encode 3x decode (support 4:2:2 color formats) 2 4:2:2 encoding and decoding 4

Error-correcting code (ECC)

None 2 Yes 18

Form factor

Form Factor

GPU memory

Graphics APIs

DirectX 12 Shader Model 6.6 OpenGL 4.6 Vulkan 1.3 3 DirectX 12 Shader Model 6.6 OpenGL 4.63 Vulkan 1.3 2 OpenGL 4.6 Vulkan 1.3 8 OpenGL 4.6 Vulkan 1.4 2 Vulkan 1.3 1

Licence required

No 20 No 20 Yes 1 Yes 1

Licence Required

No 20 No 20 Yes 1 Yes 1

Main Port Type

Ethernet 4 Fibre 17

Max simultaneous displays

2x 7680 x 4320 @ 100 Hz 5 2x 7680 x 4320 @ 30Hz 1 2x 7680 x 4320 @ 60 Hz 8 4x 5120 x 2880 @ 60 Hz 1 Up to four simultaneous displays 1

Memory bandwidth

Memory Interface

128-bit 5 160-bit 2 192-bit 3 256-bit 2 384-bit 2 512-bit 4 64-bit 1

Model

NVIDIA NVLink

No 7 None 2

NVIDIA NVLink bandwidth

No 7 None 2

Port Count

10 2 16 3 22 2 32 6 52 2 60 2 64 4

Port Speed

10/100/1000 1 100GbE 9 200Gb/s 100Gb/s 2 200GbE 4 25GbE / 100GbE 3 400GbE 2

Power connector

16-pin 1 1x PCIe CEM5 16-pin 7 PCIe 5.0 x16 1 PEX 8-pin 1

Power consumption

Power Consumption

Redundant Power Supply

No 5 Yes 16

RT Core performance

RT Cores

SFP Ports

16 3 2 1 22 1 32 6 4 2 60 2 64 4 8 2

Single-precision performance

Stackable

Both 1 No 2 Virtual 18

Standard Warranty

1 Year 20 NA 1

System interface

PCI Express 5.0 x16 1 PCIe 4.0 x 16 5 PCIe 4.0 x 8 3 PCIe 5.0 x 8 2 PCIe 5.0 x16 6 PCIe Gen4 x16 64GB/s 1 PCIe Gen4 x8 2 PCIe Gen4: 64 GB/s 1 PCIe Gen4x16: 64GB/s bi-directional 1

Tensor Cores

192 2 24 1 240 1 400 1 568 2 5th Generation 8 72 1 752 (5th Gen) 1 88 1

Tensor performance

1044.4 TFLOPS 1 1457.0 TFLOPS 1 191.9 TFLOPS 1 21.7 TFLOPS 1 306.8 TFLOPS 2 53.8 TFLOPS 1 634.0 TFLOPS 1

Thermal solution

Active 15 Passive 5

vGPU profiles supported

No 2 See the Virtual GPU Licensing Guide 2 See Virtual GPU Licensing Guide 1

vGPU software support

No 4 NVIDIA RTX Desktop Manager 1 NVIDIA Virtual PC NVIDIA Virtual Applications NVIDIA RTX VirtualWorkstation NVIDIA AI Enterprise 2 NVIDIA vPC/vApps NVIDIA RTX Virtual Workstation 2 NVIDIA vPC/vApps NVIDIA RTX vWS NVIDIA AI Enterprise 1 Supports vGPU 15.2 or later 1 Yes 1

VR ready

No 2 Yes 5

Vendors

Categories

Cloud Managed

Compute APIs

Cooling

CUDA Cores

Display connectors

Encode/decode engines

Error-correcting code (ECC)

Form factor

Form Factor

GPU memory

Graphics APIs

Licence required

Licence Required

Main Port Type

Max simultaneous displays

Memory bandwidth

Memory Interface

Model

NVIDIA NVLink

NVIDIA NVLink bandwidth

Port Count

Port Speed

Power connector

Power consumption

Power Consumption

Redundant Power Supply

RT Core performance

RT Cores

SFP Ports

Single-precision performance

Stackable

Standard Warranty

System interface

Tensor Cores

Tensor performance

Thermal solution

vGPU profiles supported

vGPU software support

VR ready

Filters and sorting

Vendors

Categories

Cloud Managed

Compute APIs

Cooling

CUDA Cores

Display connectors

Encode/decode engines

Error-correcting code (ECC)

Form factor

Form Factor

GPU memory

Graphics APIs

Licence required

Licence Required

Main Port Type

Max simultaneous displays

Memory bandwidth

Memory Interface

Model

NVIDIA NVLink

NVIDIA NVLink bandwidth

Port Count

Port Speed

Power connector

Power consumption

Power Consumption

Redundant Power Supply

RT Core performance

RT Cores

SFP Ports

Single-precision performance

Stackable

Standard Warranty

System interface

Tensor Cores

Tensor performance

Thermal solution