Friday, April 11, 2014

A Primer on Memory Storage Density

Memory storage density is a measure of the quantity of information bits that can be stored on a given length of track, area of surface, or in a given volume of a computer storage medium.  Generally, higher density is more desirable, for it allows greater volumes of data to be stored in the same physical space. Density therefore has a direct relationship to storage capacity of a given medium. Density also generally has a fairly direct effect on the performance within a particular medium, as well as price.

Examples

Hard drives store data in the magnetic polarization of small patches of the surface coating on a (normally) metal disk. The maximum areal density is defined by the size of the magnetic particles in the surface, as well as the size of the "head" used to read and write the data. The areal density of disk storage devices has increased dramatically since IBM  introduced the RAMAC, the first hard disk, in 1956. RAMAC had an areal density of 2,000 bit/in2.  Commercial hard drives in 2005 typically offer densities between 100 and 150 Gbit/in², an increase of about 75 million times over the RAMAC. In 2005 Toshiba  introduced a new hard drive using perpendicular recording, which features a density of 179 Gbit/in².  Toshiba's experimental systems have demonstrated 277 Gbit/in², and in 2006 Seagate Technology demonstrated a drive with a 421 Gbit/in² density. It is expected that perpendicular recording technology can scale to about 1 tBit/in² at its maximum.

Compact Discs (CDs), another common storage media of the early 2000s, stores data in small pits in plastic surface that is then covered with a thin layer of reflective metal. The standard defines pits that are 0.83 micrometers long and 0.5 micrometers wide, arranged in tracks spaced 1.6 micrometers apart, offering a density of about 0.90 Gbit/in².  DVD  disks are essentially a "product improved" CD, using more of the disk surface, smaller pits (0.64 micrometers), and tighter tracks (0.74 micrometers), offering a density of about 2.2 Gbit/in². Further improvements in HD DVD and Blu-ray offer densities around 7.5 Gbit/in² and 12.5 Gbit/in², respectively (for single-layer devices in both cases). When CDs were first introduced they had considerably higher densities (and overall capacity) than then-current hard drives, however hard drives have improved much more quickly than optical media, and by the time the latest blue-laser systems become widely available in 2007, the average hard drive will store somewhere between 500 and 750 GB with densities between 150 and 250 Gbit/in².

A number of technologies are attempting to surpass the densities of all of these media. IBM's Millipede memory is attempting to commercialize a system at 1 Tbit/in² in 2007 (800 Gbit/in² was demonstrated in 2005). This is about the same capacity that perpendicular hard drives are expected to "top out" at, and Millipede technology has so-far been losing the density race with hard drives. Development since mid-2006 appears to be moribund, although the latest demonstrator with 2.7 Tbit/in² seemed promising. A newer IBM technology, racetrack memory, uses an array of many small nanoscopic wires arranged in 3D, each holding numerous bits to improve density. Although exact numbers have not been mentioned, IBM news articles talk of "100 times" increases. Various holographic storage technologies are also attempting to leapfrog existing systems, but they too have been losing the race, and are estimated to offer 1 Tbit/in² as well, with about 250 GBN/in² being the best demonstrated to date- for non-quantum holography systems. Finally, molecular polymer storage has been shown to store 10 Tbit/in².

By far the densest type of memory storage experimentally to date is electronic quantum holography.  By superimposing images of different wavelengths into the same hologram, a Stanford research team was able to achieve a bit density of 35 bit/electron (approximately 3 Exabytes/in2.) This was demonstrated using electron microscopes and a copper medium as reported in the Stanford Report on January 28, 2009.

Effects on Performance

Increasing storage density of a medium almost always improves the transfer speed at which that medium can operate. This is most obvious when considering various disk-based media, where the storage elements are spread over the surface of the disk and must be physically rotated under the "head" in order to be read or written. Higher density means more data moves under the head for any given mechanical movement.

Considering the floppy disk as a basic example, we can calculate the effective transfer speed by determining how fast the bits move under the head. A standard 3½” floppy disk spins at 300 rpm, and the innermost track about 66 mm long (10.5 mm radius). At 300 rpm the linear speed of the media under the head is thus about 66 mm x 300 rpm = 19800 mm/minute, or 330 mm/s. Along that track the bits are stored at a density of 686 bit/mm, which means that the head sees 686 bit/mm x 330 mm/s = 226,380 bit/s (or 28.3 KiB/s).

Now consider an improvement to the design that doubles the density of the bits by reducing sample length and keeping the same track spacing. This would immediately result in a doubling of transfer speed because the bits would be passing under the head twice as fast. Early floppy disk interfaces were originally designed with 250 kbit/s transfer speeds in mind, and were already being outperformed with the introduction of the "high density" 1.44 MB (1,440 KiB) floppies in the 1980s. The vast majority of PCs included interfaces designed for high density drives that ran at 500 kbit/s instead. These too were completely overwhelmed by newer devices like the LS-120, which were forced to use higher-speed interfaces such as IDE.

Although the effect on performance is most obvious on rotating media, similar effects come into play even for solid-state media like Flash RAM or DRAM. In this case the performance is generally defined by the time it takes for the electrical signals to travel though the computer bus to the chips, and then through the chips to the individual "cells" used to store data (each cell holds one bit).

One defining electrical property is the resistance of the wires inside the chips. As the cell size decreases, through the improvements in semiconductor fabrication that lead to Moore’s Law, the resistance is reduced and less power is needed to operate the cells. This, in turn, means that less electrical current is needed for operation, and thus less time is needed to send the required amount of electrical charge into the system. In DRAM in particular the amount of charge that needs to be stored in a cell's capacitor also directly affects this time.

As fabrication has improved, solid-state memory has improved dramatically in terms of performance. Modern DRAM chips had operational speeds on the order of 10 ns or less. A less obvious effect is that as density improves, the number of DIMMs needed to supply any particular amount of memory decreases, which in turn means less DIMMs overall in any particular computer. This often leads to improved performance as well, as there is less bus traffic. However, this effect is generally not linear.

No comments:

Post a Comment