Intelligent docking stations: The future of hybrid devices?

Early this week I watched Ubuntu’s teaser on what they have prepared for hybrid phone/tablets/desktops and I said to myself (and I am sure that many others did as well) “this is exactly what I have been looking for, but so far no one managed to deliver”.

If you haven’t seen the video, here’s the inside scoop. Ubuntu righteously argues that a smart phone is a computing device like any other (be it a tablet or a PC), but it is limited by its form factor. So what if we could attach that phone into a tablet screen, and it becomes a tablet, or to a docking station and it becomes a PC?

This concept isn’t necessarily new. Motorola already tried this before, and so did ASUS. But the advantage here is that we are talking about Ubuntu, which is one of the richest Linux-based ecosystems out there. This is what got my attention.

There is just one caveat, though. A smartphone might have enough processing power for most general users, but this is never going to be enough for the knowledge worker that needs processing power for tasks such as research and development. Data mining, artificial neural networks, CAD/CAM, and so forth; for these disciplines, the more computing power they get the better. So only a fool would think that a simple ARM or Atom CPU is enough for everybody.

So I was thinking… With the speed we achieved in expansion buses like PCI express, and with interfaces like Thunderbolt, would it be possible to design docking stations that would do more than provide just a bunch of extra connections?

Here is what I have in mind, which would really push the edge of this concept of hybrid tablet/PC devices. Imagine a x86-64 based tablet; perhaps something with hardware specifications similar to Microsoft’s Surface Pro. Now imagine that, through a PCI express interface, you plug this tablet into a docking station that has an optical drive and extra ports such as multiple display ports, PC card, USB 3.0, eSATA, etc. That would be cool, but nothing new so far.

Now imagine that in this docking station, there is also extra RAM and additional GPU. So following a similar approach to NVIDIA’s SLI or AMD Crossfire, the additional GPU could grant extra computing power to the device. Considering NVIDIA GPUs, there is also the concept of CUDA.

Of course this is just an idea but there are challenges involved, since (at least for now) SLI and Crossfire works only to combine the same model of graphic cards; but I imagine that this wouldn’t be impossible. So perhaps instead of additional GPUs, this docking station has a bunch of co-processors, like the Intel Xeon Phi.

Then you plug your tablet device into this docking station, and the operating system then makes use of all this extra computational power. Your tablet now becomes a full-fledged workstation. Moreover, we could have different docking stations for different purposes. For instance, a musician or sound engineer could have a docking station that provides an enhanced sound system, such as a 12 channel mixing table. A trader docking station could have 6 display ports and additional features focus the business/enterprise user. Now, that is something I would look forward to!

I did a quick check online to see if there is anything similar to this idea in the works, and (perhaps unfortunately) I found out some patents that tend to overlap with this idea. And I meant perhaps unfortunately because whether those patents are going to be used to create something amazing or to just to troll innovators, is yet to be seen.