Nvidia Quadro K5000 – the new card in town for HDX 3D and HDX 3D Pro

Written by Thomas Poppelgaard on. Posted in 3DConnexion, Autodesk, Dell, HDX 3D, HDX 3D Pro, HP, IBM, Microsoft, XenApp, XenDesktop

Hi all, i am so excited to share this news with you.

I love that Citrix released in Citrix XenDesktop 5.6 FP1 a new version of HDX 3D Pro, which reduced the bandwidth consumption and increased framerate, this is a huge improvement of the technology that i adore. A lot is happening in this space and soon we will see the new Nvidia VGX card that will be able to virtualize the GPU and have multiple users on 1 GPU. NVIDIA says they will be able to have 100 users using this card. For HDX 3D Pro i would cut that down to 10-20, why cause users that normal works with a card like the Quadro 4000 are using the 256 CUDA cores for 3D rich applications like Autodesk Inventor, Dassault Catia/Solidworks, ProE + more. Again time will tell when Nvidia VGX Model 1 is released and if i know NVIDIA right, they will ship cards with more powerful GPU chips on the VGX in the future.

Nvidia have released a Quadro graphic card thats based on the new Kepler technology.
The card is called Quadro K5000.

  • This card is 1/3 the price of the Quadro 6000 and look at the numbers it delivers in the chart i have created.
  • The CUDA cores it delivers are so impressive 3x more than Quadro 6000.
  • The Power consumption is 45% less than the most powerful Quadro card.

** update 22 February 2013

If you want to use the Nvidia Quadro K5000 with GPU pass-through the card is now supported from NVIDIA.

GPU pass-through is supported with Citrix XenServer 6.x and VmWare vSphere 5.1
Above is information confirmed from NVIDIA.

Look out for the Nvidia VGX K2 thats will be sold in Q1 2013. This card is 2x K5000 on 1 card and works with GPU pass-through.

 

Quadro K5000
Quadro 4000
Quadro 5000
Quadro 6000
VGX Model 1
CUDA Cores1
1536
256
352
448
768
Single Precision Compute Performance
2.1 Teraflops
0,4 Teraflops
0,7 Teraflops
1,0 Teraflops
GPU Memory Specs
Memory Size Total
4 GB GDDR5
2 GB GDDR5
2.5 GB GDDR5
6 GB GDDR5
16 GB GDDR3
Memory Interface
256-bit
256-bit
320-bit
384-bit
?
Memory Bandwidth (GB/sec)
173 GB/s
89.6 GB/s
120 GB/s
144 GB/s
115 GB/s
Power Specs
Max Power
122 W
142 W
152 W
204 W
150 W
FormFactor
Dual Slot
Single Slot
Dual Slot
Dual Slot
Dual Slot

Source

Read more about the NVIDIA Quadro K5000 graphic card

 

Trackback from your site.

Comments (10)

  • Max

    |

    Which driver are you using for the K5000? I have XenServer 6.0.2 with PCI pass-through to a Windows 7 x64 VM. I’m using the nVidia 307.79 driver. Windows recognizes it as a K5000 card, but it shows up as a yellow icon in device manager. It shows the error “Windows has stopped this device because it has reported problems (code 43)”

    Reply

  • Thomas Poppelgaard

    |

    Which server do you have the K5000 card in ?

    How is your BIOS configured ?

    Regards
    Thomas Poppelgaard

    Reply

  • Max

    |

    Mr. Poppelgaard. Thanks for your response! I’m using an HP z820 with VT-d enabled. I have (2) K5000’s in it. I’ve used this configuration with Quadro 4000’s and 6000’s without any problems, but I’m having some power issues with the 6000’s. That’s the main reason I’m looking at the K5000’s.

    I just verified with Nvidia that the correct driver to use with the K5000’s is 307.79 (the latest), so my problem is probably not related to the driver.

    Are you using XenServer with PCI pass-through or are you running baremetal? Maybe there is a problem with the hypervisor. One interesting thing is that when you do an lspci on the hypervisor with a 6000, it knows it’s an NVIDIA Quadro 6000. With the K5000’s the hypervisor just says “VGA compatible controller: nVidia Corporation…”

    Reply

  • Thomas Poppelgaard

    |

    Hi Max
    Great i can try helping you, i am doing my best.

    Have you tried to install the K5000 on bare metal to verify it works with Citrix XenDesktop HDX 3D Pro ? or do you use it with Citrix XenApp HDX 3D or another technology purpose ?

    It sounds like there is a issue with the XenServer GPU pass-through, but i dont have any knowledge with this card. Let me try doing some research and please post your solution if you figure it out.

    Reply

  • Thomas Poppelgaard

    |

    Hi Max i just talked with NVIDIA and i can confirm that Quadro K5000 dosent work with GPU pass-through. I am investigating why this isn’t possible. The VGX K2 which have 2 Quadro K5000 on 1 board will support GPU pass-through.

    Reply

  • max

    |

    Hi Thomas, Thanks so much for your reply and speaking with NVIDIA! I also just heard the same thing from HP who contacted NVIDIA. They say that the K5000 will not work in GPU passthrough mode. I’m going to stop trying to get it to work now :-) Hopefully they’ll support it in the near future. You have a great blog!

    Reply

  • DW

    |

    Thomas,

    I’m in the middle of a GPU pass-through project supported by Nvidia. It seems like they’re not quite making this news as public as they should. My direct contact at Nvidia knew exactly what I was researching, but was not informed that the K5000 didn’t have this capability. It was only after I sent help request e-mails to him and his engineers that this lack of capability came out.

    Wrote a mini-press release on my site to talk about it: http://dwhynes.com.

    The guys at Nvidia were very apologetic for not disclosing this sooner, and are sending a 6000 for testing instead. Hopefully I can repeat the test with the VGX K2 later on.

    I’m a bit of an online lurker, but I dig your site. Cheers!

    Reply

  • Thomas Poppelgaard

    |

    Max@ Thanks for the kind words

    DW@looks interesting the project you are doing, i would like to talk with you on skype about your project maybe we could sync on the projects i am doing.

    Reply

  • Eddy Lin

    |

    Thomas,
    Do you have the source/link from Nvidia confirming this? I can’t seem to find it searching. Thanks.

    Reply

  • Thomas Poppelgaard

    |

    Hi Eddy I dont have a source / link, but i can confirm from NVIDIA and Citrix that it works with GPU pass-through now, cause they have changed this in the Nvidia driver.

    /poppelgaard

    Reply

Leave a comment

Turn on pictures to see the captcha *

Recent Comments

Danny de Wolf

|

Hey Thomas,
Great work, thanks. I’m still running into an issue that I cannot seem to resolve. I was hoping you might have an idea where I should look ?
I’ve got an application server with ES5.4, and a separate SQL server with the Reporting installed. I can get to the Reporting websites /reports and /reportserver, but when configuring Reporting within ES, it keeps saying “The report server credentials used to communicate with Reporting Services are invalid”. I tried pretty much everything I think, and cannot get my head around this.
The account that I’m using at the “Report Server Credentials” page, is a domain account, and has admin rights on the application server (for now, local admin) as well as on the Report and ES databases.
I’m a bit lost what else I can check to make this work… Any advise you can give me would be greatly appreciated :-)
Thanks.

Mike

|

+ Receiver for Windows 14.1.200
+ Profile Management 5.2

Ge

|

Same here. Version 12 and 13 of the receiver still can’t handle SHA2 so for me also a useless product.
Come on Citrix, you have implemented sha2 almost in all receivers so it should be rocket-science for you.

Dan

|

Could you tell me how much will cost this solution ?

only hard+ soft/licenses , without configuration/consultancy .

thanks,
Dan

Victor

|

Changing my playground from XD71 to XD75 and upgrading to the latest VGX is very disappointing.
I can’t see anymore the normal output for nvidia-smi command and i also lack the performance level but also the performance graphs in XenServer.
Tried to rollback and i was unsuccessful.

Feeds from Citrix Blog :