Phase-change memory

alleycat · Post by **alleycat** » Wed Dec 13, 2006 5:13 pm

Smaller, faster and more reliable than flash, phase-change memory has the potential to replace mechanical disks in many applications.

jaganath · Post by **jaganath** » Thu Dec 14, 2006 9:56 am

Solid-state hard disks can't come soon enough; all the suspension and mucking about with sorbothane will be totally unnecessary then.

darthan · Post by **darthan** » Thu Dec 14, 2006 1:29 pm

Unfortunately the many applications that phase change can replace hard disk is in things like cell phones and stuff. This type of memory won't have the capacity to replace hard drives for several years, at least.

highlandsun · Post by **highlandsun** » Sat Dec 16, 2006 2:23 pm

The phase change requires using a small charge to heat the material to above its melting point, then cooling it back down so that it either freezes into a crystalline or an amorphous form. The obvious question then is, what is the melting point of this alloy? When installed into a system, what operating temperature ranges will it tolerate before the ambient temperature starts interfering with either the heating or the cooling phase?

Since we're talking about applying heat to toggle individual bits, how densely can the bit cells be packed before the residual heat starts interfering with operations on neighboring cells, or even subsequent operations on the same cell?

jaganath · Post by **jaganath** » Sat Dec 16, 2006 2:56 pm

obvious question then is, what is the melting point of this alloy?

http://en.wikipedia.org/wiki/GeSbTe

GeSbTe, or Germanium-Antimony-Tellurium, also known as GST or 225, is a phase change material from the group of chalcogenide glasses, used in rewritable optical discs and phase-change memory...The new phase-change memories are likely using N-doped GeSbTe semiconductor. The melting point of the alloy is about 600 Â°C and the crystallization temperature is about 400 Â°C.

From that I would say there is nowhere on earth where the ambient temperature is high enough to affect this process.

Since we're talking about applying heat to toggle individual bits, how densely can the bit cells be packed before the residual heat starts interfering with operations on neighboring cells, or even subsequent operations on the same cell?

From the press release:

http://www-03.ibm.com/press/us/en/press ... /20744.wss

The device's cross-section is a minuscule 3 by 20 nanometers in size, far smaller than flash can be built today and equivalent to the industry's chip-making capabilities targeted for 2015. This new result shows that unlike flash, phase-change memory technology can improve as it gets smaller with Moore's Law advancements.

I don't know why the heat from one cell doesn't seem to affect other cells even at nanometre scales.

pyogenes · Post by **pyogenes** » Sat Dec 16, 2006 3:24 pm

sounds like more vaporware to me. Maybe I'm too jaded, but I refuse to believe something will be released until the day after it's released.

Things that were promised to be released a long time ago (working prototypes existed):

holographic hard drives. When were they annouced? 1995? 1996?

fuel cells for laptops. A far more environmentally friendly way to power your laptop (uses renewable resources unlike drawing power from your outlet which is most likely fossil fuel based). That was at least 5 years ago they promised it'd be more common.

Paper-based displays. Computer monitors made of paper and wire.

3 and 4 state memory (different groups designed them). Basically instead of memory that could store only 2 distinct values (0 or 1), it'd be able to store 3 or 4 values (I guess they'd be call trits and quits instead of bits)