The VMware PCoIP ‘Killer App’
Protocol, VDI September 2nd, 2009
With the announcement of the inclusion of the PC over IP (PCoIP) Teradici in VMware View this week at VMworld. I think that there is something people may be missing.
The big disadvantage of the original hardware to hardware PCoIP implementation was that each connection to the server required it’s own Teradici card. This is obviously not a scalable solution. As the software to software solution is unveiled at VMworld, the attention seems to be on the fact you can get the performance without stuffing your servers full of Teradici cards. To my mind the software to software approach has a big flaw, you need power on the client. Power on the client means either a full PC on the other end, which defies the point, or a really expensive thin client.
The real key would be to go from software to hardware. A software client on the server communicating with a hardware Teradici chip on the client. You could avoid all the issues of managing the ‘almost PC’ modern thin clients and go back to the cheap, minimal management, devices I think thin clients should be.
I’m curious as to why this is not being made more of as the client devices are already there like this one from Samsung and if you look at the Teradici video on Brian Maddens site they say it will work.
As the devices get cheaper, maybe down to about $200 with the great performance of PCoIP I can see this being the ‘killer app’ for VMware in this space.
September 3rd, 2009 at 7:27 pm
Hi Jim,
I completely agree the VMware View with software PCoIP to a hardware zero client makes a lot of sense. This is exactly what we are showing at VMworld this week. It will be in the video interview that Brian took a couple of days ago. Also, I will take some video myself and post to TeradiciLabs on Youtube.
Don’t forget the client antivirus… since the hardware PCoIP zero clients are PCoIP algorithms baked directly on the chip silicon. There is no need to load an anti-virus software since the hardware zero client is already secure. One less thing to update (and worry whether your clients have enough memory etc…)
Would be happy to discuss the roadmap for these devices directly with you.
Cheers,
Stu
Director of Biz Dev, Teradici
September 7th, 2009 at 6:25 am
Very interesting idea Jim.
The server-side implementation of PCoIP was always hamstrung by the requirement for a dedicated PCoIP processor for each session. Moving to a s/w based solution should open up the data center to PCoIP in a way that was simply not possible in a h/w only inplementation.
However, I’m not sure that a h/w assisted client implementation would reach killer app status. Getting the price of a conventional thin-client down to acceptable levels is still an uphill battle, adding another proprietary processor to the mix will make that battle even harder. But assuming that manufacturers can address the cost, the big concern I have when dealing with any proprietary hardware (I used to design hardware BTW) is that once it is baked it can’t be modified. Upgrading a thin-client to support the latest enhancements for ICA/HDX is easy, but if you rely on dedicated h/w to manage the PCoIP implementation you are locked in at a fixed point in time and any protocol improvements made after that point are only going to be accessible by throwing out the old thin-client and buying a new one.
I’d be far more accepting of a solution that didn’t have to have proprietary silicon to delivery acceptable performance.
Regards
Simon
October 27th, 2009 at 10:20 pm
Jim,
This is absolutely the issue I saw when viewing Teradici at the IPExpo in London. Nobody wants to talk about the double cost on hardware for the server and client. And it’s a huge implementation hurdle. It seems like the hardware zero client may go a ways to fixing that issue.
However, the bigger elephant in the room that I’ve discovered is bandwidth utilisation. I’ve been told that it can be throttled on the client end.
Perhaps Stu can comment on this?