I'm looking at the photos you took- I'm pretty sure your laptop has a standard PCMCIA/Cardbus/PC-Card port and you've got an Expresscard/34 device docked in a to PCMCIA-Expresscard adapter.
http://en.wikipedia.org/wiki/ExpressCard has pretty good images of Expresscard 54 and 34 cards compared to Cardbus/PCMCIA/PC-Card (whichever one you want to call it).
PCMCIA has access to a 32bit PCI system bus, for something like 133.3 MB/sec bandwidth. Expresscard 1.0 has access to at least a 1x PCIe lane, for a maximum of bandwidth of something like 300 MB/sec. Expresscard 1.0 can also access the USB bus, for a bandwidth of 60 MB/sec. Expresscard 2.0 improves things quite a bit, but it's not on the market yet. Here's a table I just found
http://en.wikipedia.org/wiki/List_of_device_bandwidths#Portable
So the big picture of why I'm wondering this at all goes back to the ViDock 2. I'm thinking about buying one so that I can phase out my desktop and just use my tablet (4500MHD graphics are pretty piss poor). It's retails for almost $400 and is hard to find below that because there's only one north American distributor (the promotional price they've got now is actually kind of tempting...). I'm hesitant to blow that much money on what is basically an external GPU / USB hub combo, but if it'll last me a while, I might pick one up. The ViDock 2 is Expresscard 2.0 compatible, so it could potentially last very long.
It's complicated though, because I would want to use it as both an external GPU and a USB hub. It's not a docking station, but it's pretty close. I could bring my tablet home, plug in power, and slide the ViDock adapter into the Expresscard slot for all my peripherals. But especially now, if Expresscards cannot simultaneously access both the PCIe and USB buses, then I've got a problem. Because then for my use, everything I plug into the USB ports is going to take away from what little video bandwidth there is to begin with.
Most current midrange GPUs are all on PCIe x16. Some might be x8, I don't know. The point is, bus bandwidth isn't really limiting a GPU at x16 yet. But when you dial it down to a 1x lane (like a ViDock must do in Expresscard 1.0 mode), bandwidth starts to become a limiting factor.
http://www.tomshardware.com/reviews...tml?xtmc=pcie_express_scaling_analysis&xtcr=1 Tom's Hardware did an analysis of just that a long time ago with some old GPUs. They only compared an 8800GTS vs an X1900XTX. Not quite current generation, but pretty close. The nVidia pretty much needs all 16 lanes. You see FPS drop proportionally to PCIe lanes. The AMD is another story. It gives about 80% performance at PCIe x1, and the differences between x4, x8, and x16 performance are minimal.
So if Expresscard can only run one of the two buses at once, then using the ViDock's hub function directly takes away from video performance and I probably won't be interested until Expresscard 2.0 is prevalent on laptops. If, however, Expresscard 1.0 can access both buses simultaneously, then I can use a ViDock happily, today, as a pseudo docking station. I'd also know that later on down the road, I can upgrade the GPU and stick it the ViDock in an Expresscard 2.0 capable laptop, and be good to go with USB 3.0 for the hub and PCIe x2 for the GPU.