GPUs EUC Datacenter comparison & reviews

Summary
Rank
2nd 1st 3rd
Score
0%
0%
User Reviews
Question?
Not Enabled Not Enabled Not Enabled
Analysis expand by Team RGE by Team RGE by Team RGE
Select All
General expand
0%
0%
0%
  • Fully Supported
  • Limitation
  • Not Supported
  • Information Only
Pros
  • + 1st Hardware based virtualized GPU solution, using SR-IOV FLR (Function Level Reset) preventing rogue guest from crashing the entire system.
  • + Predictable performance
  • + The GPU is designed for high performance graphical environments.
  • + Good performance in single user scenario, and multi-user XenApp scenario.
  • + Proven technology in consumer market.
Cons
  • - Documentation and information is dated.
  • - Slow adoption of (new) OEMs, Guest OS and Virtual Infrastructure platforms.
  • - No community and limited visibility in customer adoption.
  • - Solution (Hardware + Software) is perceived as expensive.
  • - Limited use-case, good performance in single user scenario, Limited server/OEM support.
  • - Limited Virtual Client Computing support.
  • - No community, limited information.
  • - Dated APU (2015)
  Content  
  •  
Content creator
This content is created by TeamRGE members: Patrick van den Born, Jits Langedijk and Ruben Spruijt
This content is created by TeamRGE members: Patrick van den Born, Jits Langedijk and Ruben Spruijt
This content is created by TeamRGE members: Patrick van den Born, Jits Langedijk and Ruben Spruijt
  •  
Vendor description
NVIDIA is an American technology company based in Santa Clara, California. It delivers the engines for todays Artificial Technology. NVIDIA designs graphics processing units (GPUs) for a variety of solutions, e.g. AI, Gaming, Cryptocurrency, and professional markets, as well as system on a chip units (SoCs) for the mobile computing and automotive market. Since 2014 NVIDIA has shifted to become a platform company focused on five markets – AI, Gaming, Professional Visualization, Datacenters and Automotive. 
  •  
Product overview
The AMD FirePro family of hardware-virtualized GPUs with Multiuser GPU (MxGPU) technology delivers a simple, secure and cost effective solution for high performance virtual desktops and workstations.
All the features and functions are based on NVIDIA GRID 5.0. NVIDIA GRID extends the power of GPU technology to deliver virtual applications, desktops and workstations for every user.
  •  
Community Specialist
TeamRGE (Patrick, Jits, Ruben)
TeamRGE (Patrick, Jits, Ruben)
TeamRGE (Patrick, Jits, Ruben)
  Support Status  
  •  
Product is supported (Not End Of Life )
  •  
Long-term branch maintenance
Yes
  Maturity  
  •  
Introduction date
  •  
Proven Technology
  •  
Public enterprise references in Europe
No
No
  •  
Public enterprise references in US
No
No
  •  
Large community resource around GPU Virtualization - Datacenter
  Solution Pricing  
  •  
Hardware
Contact vendor/OEM
Contact vendor/OEM
  •  
Software
No
No
  •  
Support
N/A
N/A
  •  
Education pricing
N/A
N/A
  •  
MSP pricing
N/A
N/A
  •  
Named User
N/A
N/A
  •  
Concurrent User
N/A
Yes
N/A
  •  
VM based
N/A
N/A
  •  
Host based
N/A
N/A
  •  
GPU board based
N/A
N/A
  •  
Free Trial software
Yes
N/A
Hardware expand
0%
0%
0%
  Product Specification  
  •  
GPU Model
  •  
GPUs per board
  •  
Transistors
  •  
Total board power (W)
  •  
GPU Clock - Base (MHz)
800MHz
  •  
GPU Clock - Idle (MHz)
  •  
NVIDIA CUDA cores
N/A
  •  
Max. users
32
  •  
PCI Express interface
PCI Express Gen3 x16 system interface
PCI Express Gen3 x16 system interface. Lane and polarity reversal supported
  •  
Board design
  •  
Power connectors
1 x CPU 8-pin auxiliary power connector, 1 x  2-pin power brake header
  •  
Thermal Solution
Active/Passive Cooling
  Memory Specifications  
  •  
Memory Clocks - Performance (MHz)
  •  
Memory Clocks - Idle (MHz)
  •  
Memory Size
16 GB (8 GB per GPU)
  •  
Memory I/O
  •  
Memory configuration
GDDR5 - 16GB (2 x 8GB)
32 pcs 256M × 16 GDDR5 
  •  
Memory bandwidth
  Thermal Specifications  
  •  
GPU shutdown temperature (°C)
  •  
GPU slowdown temperature (°C)
  Performance  
  •  
Pixel Fill rate
  •  
Texture Fill rate
  •  
SPFP Performance
3,763 GFLOPS x2 (7.54)
  •  
Performance per W
  Encoders  
  •  
H.264 - supported
  •  
H.264 - amount per interface board
36
  •  
H.265 - supported
Yes
  •  
H.265 - amount per interface board
  Decoders  
  •  
H.264 - amount per interface board
  •  
H.265 - amount per interface board
  Server vendor support  
  •  
Hardware Compatibility List
Yes
Software expand
0%
0%
0%
  Hypervisor support  
  •  
Microsoft Hyper-V
Yes
  •  
Citrix XenServer
No
Yes
  •  
Nutanix AHV
Yes
  •  
KVM
No
Yes
  •  
Huawei UVP
Yes
  •  
VMware ESX
Yes
Yes
  Monitoring GPU performance  
  •  
Integration with Citrix XenCenter
No
  •  
integration with VMware vCenter
Yes
  •  
Integration with Nutanix PRISM
  •  
Public SDK/API available
  Virtual Client Computing  
  •  
Citrix XenDesktop
No
Yes
  •  
Citrix XenApp
Yes
Yes
  •  
Microsoft RDSH
Yes
  •  
Microsoft RDVH
Yes
  •  
VMware Horizon
Yes
Yes
Yes
  Guest OS  
  •  
Microsoft Windows 7
Yes
Yes
  •  
Microsoft Windows 8
No
Yes
  •  
Microsoft Windows 10 - 1511
Yes
Yes
  •  
Microsoft Windows 10 - 1607
Yes
Yes
  •  
Microsoft Windows 10 - 1703
Yes
Yes
  •  
Windows Server 2012R2
Yes
  •  
Windows Server 2016
Yes
  •  
Linux - Ubuntu
Yes
  •  
Linux - SLED/SLES
Yes
  •  
Linux - Red Hat Enterprise
Yes
  API support  
  •  
DirectX 12 (2015)
  •  
DirectX 11.2 (2013)
  •  
DirectX 11.1 (2012)
  •  
DirectX 10 or lower
  •  
OpenGL 4.6 (2017)
  •  
OpenGL 4.5 (2014)
  •  
OpenGL 4.4 (2013)
  •  
OpenGL 4.3 (2012)
  •  
OpenGL 4.2 or lower
  •  
OpenCL 2.0
  •  
Shader Model 5.0
  •  
Vulkan 1.0
  •  
CUDA
Yes
  •  
OpenACC
Various expand
0%
0%
0%
  GPU functions  
  •  
Pass-Through support
Yes
  •  
BareMetal Support
Yes
  •  
Dedicated GPU (cores)
Yes
  •  
Dedicated GPU (Frame buffer)
Yes
  •  
Shared GPU (GPU cores)
  •  
Max. Frame buffer (vRAM) per VM
8 GB
  •  
Min Frame buffer (vRAM) per VM
512 MB
  •  
Graphics Mode
  •  
Compute mode
Yes
  •  
Max. density per GPU
16
16
  •  
Max. density per GPU board
32
  •  
Anti-aliasing
Yes
  Display  
  •  
Max. amount of virtual displays
4
  •  
Max. resolution per display
4096x2160 (4K DCI)
  •  
Max. vGPUs per board
32
N/A
  •  
Max. vGPUs per GPU
16
N/A
Applications expand
0%
0%
0%
  Use-case  
  •  
Office Workers (LOB, Office, browsers)
  •  
Power User - Professional Graphics
Yes
  •  
Power User - Designer
  Certification  
  •  
Adobe
No
Yes
No
  •  
Ansys
No
Yes
No
  •  
Autodesk
No
Yes
No
  •  
Bently
No
Yes
No
  •  
ESRI
No
Yes
No
  •  
HIS
No
Yes
No
  •  
PTC
No
Yes
No
  •  
Siemens PLM
No
Yes
No
  •  
Solidworks
No
Yes
No
  •  
Synerscope
No
Yes
No
  •  
Quadro support
Yes

Matrix Score

  •  
  •  
  • AMD
  • NVIDIA
  • Intel
  •  
  • 2 nd
  • 1 st
  • 3 rd
X
Login to access your personal profile

Forgot your Password?
X
Signup with linkedin
X

Registered, but not activated? click here (resend activation link)

Login to access your personal profile

Receive new comparison alerts

Show me as community member

I agree to your Terms of services

GDPR