well, i think is a very good idea on releasing two products that share hardware.
i mean, my bet was that GPD-XD2 was the same hardware than GPD-WIN, but with other features such 6" screen, 360š screen rotation and NO keyboard.
GPD-WIN will offer a keyboard that become as standard UMPC (micro-notebook)
if both of them share the same hardware, then GPD can release 3 flavors of operating systems that can be flashed on both devices:
Windows pure firmware
Android pure firmware
DualOS (Android & Windows)
this way, users can choice which device prefer depends on their needs, and which operating system want to install on it.
for GPD, this is a very easy way to update software. If they want to release a firmware with Android 6.0, users can install on any of the devices (Win or XD2).
for me, this is the best way to improve their current GPD-XD device
now,GPD will keep using ARM on their GPD-XD2. ok, nice, but what you get with this improved hardware? (assuming is the RK3399)
Better preformance, ok.
Better GPU, ok.
is there any software (except the NVidia specific apps) that you cannot run on RK3288 (GPD-XD as example)?
is there any software that RK3288 power is not enough, but RK3399 can run now?
I really donīt see a real improvement on using RK3399 on the new device. The design is a good improvement, ok, but the power will be wasted really.
the new SoC is an octacore? great!, but for emulation, most of emulators only uses 2 cores... (and there are some of them that are mono-core)
Donīt misundestand to me. Of course, everyone want better devices, better speed, etc. but currently android gaming evolution is stopped from a while, and there is no new releases that requires a better machine for running.
For all these reasons, i think a x86 SoC such the X8550 of GPD-Win is a very good implementation, because you get the same and enough power for running almost any android game or program, and you get the possibility to run another OS such Windows, or Linux Distro, etc. something that RK3399 cannot offer (almost for now)