I belive it, but the article mentions that the tech requires non-silicon semiconductors (enormously expensive to scale right now, think of how expensive building current-gen fabs is, and tack on the cost of experimental tech) not to mention the fact that changing the memory architecture to merge RAM and long-term storage would require significantly altered CPU design, (and probably significant OS/Kernel changes too!)
TL;DR assuming that this is real and not wierd shareholder fluff like other commenters assume, we probably won’t see it in consumer hardware (or even enterprise stuff) for a long time
yeah, ferroelectric RAM (F-RAM) has been a thing for a while. it’s persistent (nonvolatile), faster than DRAM, and more write endurant than flash. it’s not some top secret government technology either – you can just go buy it on digikey right now. the problem is that 512KB F-RAM modules are $20 a chip and ain’t nobody building a computer out of THAT
I belive it, but the article mentions that the tech requires non-silicon semiconductors (enormously expensive to scale right now, think of how expensive building current-gen fabs is, and tack on the cost of experimental tech) not to mention the fact that changing the memory architecture to merge RAM and long-term storage would require significantly altered CPU design, (and probably significant OS/Kernel changes too!)
TL;DR assuming that this is real and not wierd shareholder fluff like other commenters assume, we probably won’t see it in consumer hardware (or even enterprise stuff) for a long time
yeah, ferroelectric RAM (F-RAM) has been a thing for a while. it’s persistent (nonvolatile), faster than DRAM, and more write endurant than flash. it’s not some top secret government technology either – you can just go buy it on digikey right now. the problem is that 512KB F-RAM modules are $20 a chip and ain’t nobody building a computer out of THAT