The USB 3.0 stack is basicaly ready and on some systems it does work. However on *some* Intel chipsets the USBX.SYS host controller driver does not work yet. This is why the USB 3.0 driver has not yet been released. David Azarewicz is working on the issue to try and resolve this.
I have tested the USBXHCD.SYS driver on a few machines, with "interesting" results (completely unscientific):
Lenovo ThinkPad L530, with Intel USB 3 adapter. The driver loads, but it doesn't recognize when a USB device is attached. That may have something to do with the internal USB HUB since one of the USB 2 ports doesn't work either, when the USB 3 driver is loaded.
Lenovo ThinkPad T510, with a StarTech ExpressCard USB 3 adapter (Renesas chipset, apparently). I can read from a USB 3 stick, but writing has problems.
Lenovo ThinkPad L530, with the StarTech ExpressCard. Won't even POST. Obviously some sort of conflict with the internal USB 3 adapter (nothing, at all, to do with any operating system).
IBM ThinkPad T43, with the StarTech ExpressCard. A quick test shows that it works as expected. It seems to be not much faster than USB 2, but that is likely a limitation of the machine speed. I have not stress tested this one, yet.
Asus M3A78-EM motherboard, with PCIe USB 3 adapter (Fresco Logic). The driver loads, but nothing happens when I plug in a USB device.
I have not, yet, tried my Asus A88XM-A machine. It has an AMD adapter, which seems to use the Fresco Logic driver, in windows.
So, three machines (ThinkPads), using the exact same ExpressCard adapter, and USB stick, have different indications. One of them actually works.
This sort of thing has got to drive programmers up the wall.
The big question is, does that mean it works on all AMD chipsets or have they been forgotten?
AMD motherboards may use different chipsets, it depends on which chipset is cheapest when the board is built. My Asus A88XM-A calls it an AMD device, but windows uses the Fresco Logic driver. In theory, ALL USB 3 adapters, and devices, are supposed to conform to the same set of rules. It appears that manufacturers are still ignoring the rules, as they always did with earlier versions of USB (including Intel, and they, apparently, wrote the rules).