I'm curious about the growing crop of ultra low-end laptops, tablets, and convertibles which use eMMC for the system drive. My understanding is that this is basically a more mature type of flash storage - essentially an SD card soldered to the motherboard - interfaced in such a way as to make the OS see it as a fixed drive. This approach seemed to work well on smaller devices like smartphones that are not constantly writing to the system drive ... but Windows?

Background: I have long been enamored with portable applications (PortableApps) - software that can run without installation, and leaving little-to-no trace on the host system. The prototype use of this type of software is to run the apps directly from a USB flash drive. The traditional caution is to reduce operations that write to that flash drive. Too many write operations on a slow storage device will bog things down, and will eventually slow down and wear out the flash drive. So if you're using a portable web browser, for instance, you want to turn off the web page cache so that things run totally in RAM.

Similarly if you're running an OS on a flash drive (e.g. Windows To Go ... or anything else [Linux, etc]) you also want to minimize write operations to the booted drive. Turn off the pagefile, scheduled defrag, and so forth.

Which brings me back to this whole idea of running Windows on a laptop which uses eMMC as the system drive. Bearing in mind everything above - how do these actually perform, and how do they get away with it? Windows is CONSTANTLY writing to the system drive: registry changes, logs, system events, pagefile (indeed, on a machine having only 2GB RAM) - you name it.