Feasability and roadmap query

Hi there,

I have been reading up on USB dev boards, and wanted to ask a few questions before I commit to anything.

First, do you see a high-speed capable version of teensy anytime soon?

Second, I am investigating constructing my own custom keyboard a mouse devices, to work such that all advanced and programmable features should be fully programmable on and through the device itself, such that in most cases, no further drivers would be needed by the host OS. This means extra memory, touch screen, gui and my own wiring, etc. Would the teensy 3.1 be able to handle all the extra devices attached simultaneously?

Finally, I am an assembly-language-only programmer. I want to be able to create my own binaries using my own assembler, and put them onto a teensy whilst retaining the ability of the teensy to identify itself to the host machine as whichever device is required. How difficult is this likely to be? (I've done plenty of CPU programming, but never from this side of the bus connection and I want to gauge the situation before I commit wholesale). To which memory address will the bootloader and proprietry code load my programs? I don't want to nuke anything I may have to rely on later.

Thanks for your time
Patrick

If you mean the USB 2.0 high-speed data rate of 480 Mbit/s, it doesn't seem very likely with an ARM CPU running at 48 or 72 MHz.

Assembly language is certainly doable. But you would proceed at a snail's pace. A very, very long learning curve. Difficult to easily re-use the vast amount of library module code in C/C++.
Doing ARM applications in assembly would deliver you into a 0.01% demographic.

If your joy is assembly, then you probably need to focus on very small memory 8 bit MCUs. Especially if you must use your own assembler and linker.