Thin and Zero Clients with ProE users to avoid WAN and Windchill Replication Issues.

[patrickchin2000]

The subject is pretty much the solution. Why move mountains of data through small costly WAN pipe lines? Just view it and send keyboard and mouse inputs.

[GaryMansell]

Patrick,

This is an area of interest for me - do you have any details that you might like to share relating to the solution?

Rgds

Gary

[patrickchin2000]

Zero or Thin clients are now improving with VM improving capabilities with virtualized GPU:

• http://technet.microsoft.com/en-us/virtualization/hh278966.aspx


• http://complaintsincorporated.com/2011/02/21/virtual-gpus-monolithic-hypervisors-slim-hypervisors-oh-my/


• http://virtualizationreview.com/blogs/virtual-insider/2011/05/configure-remotefx-hyper-v-r2-sp1.aspx

Some companies have already using this technology for CAD integrations to PLM systems. Why move a mountains of data outside the server room. The whole architecture of GPU was to have a dedicated CPU with RAM to handle the graphics rendering. That technology started with single cores CPU. Now with multi-core (6/8/16),PC3 RAMand 64bit OS, you now have the abiltiy to utilize aVM host server to simulatevirutalize graphical workstations.

With these multi-core CPU and Windchill single point of failures of background methodservers, if apache and tomcat is not an issue, do you really need to cluster or even replicate. Possitives:

• Gigs to terabytes of data don't have to be transferred/replicated continuously between sites.


• Workstation crashes or loss of information is a thing of the past.


• Data integrity iscontrolled and managed in the SAN. Clientwith fastest HDsbecause of temporary space and vaulted in PLM for archival storage.


• IT managed and supported with ease


• etc


• Completely remove WAN issues


• Completely remove complexity of distributed topology of enterprise systems (less clustering and replication of applications and SANs respectfully)

Really no need for multi-monitors if youhave a 28 in or 32 inchmonitor, but there arezero or thin clients that support multi-monitors:

Zero:

• http://www.fujitsu.com/fts/products/computing/pc/zero-clients/advanced/DZ22-2/index.html


• http://www8.hp.com/us/en/hp-information/thin-client-solutions/zero-clients.html


• http://www.dell.com/us/enterprise/p/dell-fx100/pd

If virutal GPUs is still an issue, then thin clients with multi-monitor capability:

• http://www.hp.com/canada/products/landing/thin_clients/high-performance-environments.html


• http://www.dell.com/us/enterprise/p/wyse-z-class/pd


• http://www.matrox.com/graphics/en/press/releases/2012/thin_computing/epica_series_graphics_cards/wyse_z90de7_thin_client/six_monitors/


• http://myvirtualcloud.net/?p=3713

[GaryMansell]

Thanks for your links and information - I must say I was hoping to read more about your actual implementation and real world experiences...

I have been following this technology for a while now with great interest as it could have great operational and data security benefits and is very scalable, but I still feel that it is a while off for all but the largest adopters and those with the deepest pockets.

Saying that, I know of at least one of the big Automotive OEM's who are looking to deploy 3D CAD in this manner soon and they will also be using it to host suppliers using VDI 3D CAD over WAN connections too, so that no data leaves their enterprise.

The main trouble as I see it is that the technology is extremely latency sensitive and hence best works within the same country and geographical continent, but not very well to continents around the other side of the world. When you are looking to put in a global solution, this is an issue - you need to be able to virtualise your entire PLM and 3D CAD system or none at all.

In my opinion, I still think it will be a few years before it hits the mainstream, but bring it on...

Rgds

Gary