Jump to content

Programmable metallization cell

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by Maury Markowitz (talk | contribs) at 18:44, 29 October 2007 (new article). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.
(diff) ← Previous revision | Latest revision (diff) | Newer revision → (diff)

The programmable metallization cell, or PMC, is a new form of non-volatile computer memory being developed at Arizona State University and its spinoff, Axon Technologies. PMC is one of a number of technologies that are being developed to replace the widely used Flash RAM, providing a combination of longer lifetimes, lower power, and better memory density. Infineon, who licensed the technology in 2004, refers to it as conductive-bridging RAM, or CBRAM.

Description

PMC is based on the physical re-location of metallic ions within a glassy solid electrolyte. A PMC memory cell is made of two solid metal electrodes, one tungston the other silver or (more recently) copper, with a thin film of the electrolyte between them, along with a control transistor. Additional metal ions are deposited within the electrolyte.

When a negative bias is applied to the tungston electrode, a current forms within the electrolyte in a thin filament running to the opposite electrode. Ions in the electrolyte, as well as some from the now-positive electrode, are attracted to the current flow and migrate towards the filament. After a short period of time the ions flowing into the filament form a small metalic "nanowire" between the two electrodes. This dramatically reduces the resistance along that path, which can be measured to indicate that the "writing" process is complete.

Reading the cell simply requires the control transistor to be switched on, and a small voltage applied across the cell. If the nanowire is in place in that cell, the resistance will be low, leading to higher current, and that is read as a "1". If there is no nanowire in the cell, the resistance is higher, leading to low current, and is read as a "0".

Erasing the cell is identical to writing, but uses a positive bias. The copper ions will migrate away from the current, back into the electrolyte, and eventually to the negatively-charged copper electrode. This breaks the nanowire and increases the resistance again.

PMC is not the only application of this basic concept, which is becoming known as "nanoionics". Other prospective applications include dynamically-reroutable electronics, optical switches, and microfluidic valves.[1]

Comparison

The primary form of solid-state non-volatile memory in use to day is Flash RAM, which is finding use in most roles that used to be filled by hard drives. Flash, however, has a number of problems that have led to many efforts to introduce products to replace it.

Flash is based on the floating gate concept, essentially a modified transitor. Conventional transistors have three connections, the emitter, collector and base. The base is the essential component of the transistor, switching the resistance between the emitter and collector. In the floating gate transistor, the base is attached to a layer that traps electons, leaving it switched on (or off) for extended periods of time. The floating gate can be re-written by passing a large current through the emitter-collector circuit.

It is this large current that is Flash's primary drawback, and for a number of reasons. For one, each application of the current physically degrades the cell, and they will eventually not be able to be written to. Write cycles on the order of 105 to 106 are typical, limiting its application to roles where constant writing is not common. The current also requires an external circuit to generate, using a system known as a charge pump. The pump requires a fairly lengthy charging processes so writing is much slower than reading, and requires much more power as well. Flash is thus an "asymmetrical" system, much more so than conventional RAM or hard drives.

PMC, on the other hand, writes with relatively low power and high speeds. The speed is inversely related to the power applied (to a point, there are mechanical limits), so the performance can be tuned for different roles. Additionally, the writing process is "almost infinitely reversible",[2] making PCM much more universally applicable than Flash.

Another problem with Flash is that the floating gate suffers leakage that slowly releases the charge. This is countered through the use of powerful insulators surrounding it, but these require a certain physical size in order to be useful. It also requires a specific physical layout, which is different than the more tupical As Flash scales rapidly downward in size the charge leakage increasingly becomes a problem, and has led to several predictions of Flash's ultimate demise. However, massive market investment has driven development of Flash at rates in excess of Moore's Law, and startup fabs using 30 nm processes are currently (late 2007) being brought online.

PMC, in theory, can scale to sizes much smaller than Flash, theoretically as small as a few ion widths wide. It is also much simpler in layout than Flash, which should lead to simpler construction and lower costs.[2] Whether or not these advantages can be brought to market remains to be seen; the wide variety of other "Flash killers" have so far always been behind the technology curve of Flash's massive investment. However, as the CEO of one licensee claimed, "No other technology can deliver the orders-of-magnitude improvement in power, performance and cost that this memory can."[2]

Current status

Early experimental PMC systems were based on silver-doped germanium selenide glasses, but these materials were not able to withstand the temperatures used in standard CMOS fabs. Work then turned to silver-doped germanium sulfide electrolytes,[3] and then finally to the current copper-doped germanium sulfide electrolytes.

Axon has been licensing the basic concept since its formation in 2001. The first licencee was Micron Technology, who started work with PCM in 2002.[4] Infineon followed in 2004,[5], and a number of smaller companies have since joined as well.

References