I fully expect memristor-based drives to fall in price much faster than NAND-based flash drives did. They're somewhat easier to manufacture, should scale much more easily (some prototype memristor elements have already been tested around 1nm) and can actually be stacked on top of each other, unlike flash chips. This stacking ability should also increase memory densities quickly once drive manufacturers actually start developing their ReRAM chips with stacked memristor layers. Another benefit of the tech is that it produces little heat in comparison to transistor-based tech, which helps to allow for stacking, and given that memristors, at least ones similar to HP's version of the tech, utilize very little power, requiring milliamps and millivolts compared to the (if I remember right) 20V needed to flash NAND memory. I really think one thing that might hold back widespread adoption of the tech as a replacement for the flash memory business is that all these companies have just sunk billions into development and deployment of NAND flash SSD drives, and are likely going to resist scrapping the tech just as it really becomes profitable -- and paying back all the R&D funds that they spent.