Memories are one of the most important and essential components in computer systems. Whereas in these systems the main issue is loading data and processing them, there is a needed place to load this information on it: Memory.

So to increase the speed of loading and processing this information, thus increasing total speed of computer systems, the speed of loading process and processors should be increased. This is the reason of increasing the speed and frequency of processors. But what about loading issues? Loading places are memories so the speed of memories should be increased like processors.

Many different technologies are used to enhance this speed. Double Data Rate, Dual and triple channel memories, etc. all are some examples of these techniques. But even by using these technologies, the most important factor in this issue is frequency. It means however using highest frequency is needed and we can not neglect this improvement.

Nowadays available semiconductor devices are based on CMOS technology, and used material is silicon. Used memories in computer systems are based on MOSFETs which are fabricated using CMOS technology and silicon. RAM, Flash memories, recently SSDs and etc all of them are made of CMOS NAND based circuits. For example using NAND based circuits in SSDs have made them as a very high speed drives which have a very bright future. Even experts believe that in early future current HDDs will be replaced by SSDs.

On the other hand, silicon has a limitation in increasing frequency. Germanium and its compounds are known as a faster semiconductor. But making CMOS technology is not possible using these materials. Because silicon dioxide has some unique features which other materials don’t have. So stop using Germanium?!

No! Scientists are working on germanium based devices. For sure in the future we can see germanium based memories. But the technology will be different to CMOS. For example Resonant Tunneling Diode is a very high frequency device. The speed of this device is not comparable to CMOS transistors. Despite CMOS transistors work in orders of Gigahertz, RTD works in orders of Terahertz!

As frequency increases, the device size should decrease too. But while device dimension decreases to a few nanometers, then electron behavior in this device differs. Electron behavior should be calculated using quantum equations not classic ones like MOSFETs or others. This phenomenon happens in RTD, and this is the reason of naming this device as a quantum device.

Memories fabricated using RTD are very faster and smaller than regular memories. Nowadays fabrication technology is 35 nm at least. So a Transistors size should be more than 70 nm. But RTD size is just 5 or 6 nm or something like this. Even if fabrication devices improve, RTDs can be fabricated smaller than this. And we should consider: smaller means faster!

So quantum based memories will catch computer world in the future and an evolution is going to occur in computer industries. Flash memories, RAM, SSDs, and even processors and used caches in them will change. Even processor itself will be changed, silicon goes and germanium comes… Quantum world, small and fast devices,… as small as some electron, and as fast as … what you think?