Dear Nvidia GRID PROs,
this year we decided to build our own VDI infrastructure and start with a hybrid version of Hyper-V and Xen virtualization services. We’d also like to deploy a hybrid Azure Pack solution on our infrastructure.
Our main objectives are to provide services to:
- small-medium sized companies (task/knowledge users with Windows VPCs);
- small-big sized companies with 3D computing RDP demands (Adobe, Autodesk/CAD, CATIA etc.);
- healthcare and medical centers (image processing and analyzing);
We’re gonna use 14xDELL Poweredge R730s and 2xPowerVault MD3820fs with DELL 4000series/Cisco Nexus FC switches and F5 security appliances. Is it advisable for us to deploy the K1/K2/TESLA cards in a separate rack appliance (like the DELL PowerEdge C4130) or should we just place all the GPUs into the R730s?
Please also advise us which specific GPU types we should choose while building up this infrastructure (and how many of them)!
We’re expecting to have around 2000 users (Win8 with office MS apps) and 100 power/designer users with 3D compution demands like Autocad Inventor. We are looking for the best scaleable solution which is also cost effective (servers we buy cheap, the GPU cards however we cannot get discounted prices for at our supplier). That’s why we have to think twice and chosse wisely the GPU types with the best cost-value rating.
Any guidance and advice you can share with us would be greatly appreciated! Thank you very much for your help!