A CDE Definition
(NanoRAM) A non-volatile, random access memory technology from Nantero, Inc., Woburn, MA (www.nantero.com) that is being developed to initially replace flash memory and ultimately DRAM and SRAM memories. NRAM uses carbon nanotubes for the bit cells, and the 0 or 1 is determined by the tube's physical state: up with high resistance, or down and grounded.
NRAM is expected to be faster and denser than DRAM and also very scalable; able to handle 5 nm bit cells whenever CMOS fabrication advances to that level. It is also very stable in its 0 or 1 state. Using standard CMOS fabrication facilities, Nantero licensed NRAM to companies for commercial production. See future memory chips.
Spin Coat the Tubes onto the Wafer
Making NRAM cells is quite fascinating. A solution of purified carbon nanotubes is placed onto a wafer with predefined round electrodes and spun at centrifugal force to spread the fluid. The tubes wind up in random polarizations, spread evenly across the wafer and over the electrodes. Subsequent steps remove the extraneous tubes and add the interconnects. Each electrode with its carbon tubes becomes a memory cell (see below).
The Bits Are Nanotubes
Making the NRAM Cells
future memory chips
The current goal is to create a memory chip that holds its content without power, but has the speed and random access capability of dynamic RAM (DRAM) and static RAM (SRAM), the main memory and cache memory chips in today's computers. Such future chips are expected to initially provide an alternative to flash memory and eventually replace DRAM and SRAM, and ultimately the hard disk.
The flash memory in common use in camera cards, USB drives and the like is NAND flash. It is non-volatile but must be written in blocks similar to disk sectors, which makes it slower. DRAM and SRAM are fast and can be written one byte at a time, but they lose their content without power. Both flash and RAM chips (DRAM and SRAM) have drawbacks the industry would love to eliminate. In the meantime, the life of flash memory is also being extended (see charge trap flash).
An Eventual Paradigm Shift
When a storage technology with the random access and speed of RAM memory chips (see byte addressable) combined with the permanence and low cost of a hard drive merge and become the norm, there will be a major change in the architecture of software.
All operating systems and applications are designed to continuously save newly created data to storage. When main memory finally "remembers" like it did in the old days of core memory (why it was originally called "memory"), many data elements will reside in memory at all times. In addition, computers will be "instant-on" and not waste AC or battery power when idle.
Intel's 3D XPoint is a commercial product that is a step in the right direction (see 3D XPoint). Other memory technologies in various stages of development are phase change memory, programmable metallization cell, MRAM, F-RAM, NRAM and memristor.
Chips Vs. Disks - Magic Vs. Jackhammers
When one compares the internal magic of any solid state memory chip with a hard disk, it becomes very apparent that although disks are highly reliable, they represent an antiquated architecture compared to the chip. Hard disks are mechanical. An access arm moves back and forth like a jackhammer over spinning platters. Contrast this jackhammer and turntable architecture with the electron movement inside a memory chip or solid state drive (SSD).
Mobile devices use solid state drives (SSDs), and increasingly, so do laptop and desktop computers. In datacenters, SSDs are replacing banks of hard drives. There will come a time when all mechanical storage is history. See solid state drive.
Before/After Your Search Term
Terms By Topic
Click any of the following categories for a list of fundamental terms.