Matters of State:
How PCM Works, and How it Improves Performance
Today’s advanced embedded applications require greater volumes of code and data than ever before, challenging the architectural constraints of traditional memory technologies. In the absence of a single memory type that can function across multiple platforms, designers have been forced to configure increasingly complex memory subsystems to overcome these limitations. But new phase change memory (PCM) is revolutionizing the way subsystems are arranged, providing an all-encompassing technology that can meet designers’ diverse needs in a single memory device.
Featured PCM Videos
To view additional PCM videos, click on the monitor icon within the video player frame.
PCM works by taking advantage of the alterable physical and electrical properties of certain materials in order to store information on a memory device, similar to the way that high-volume rewritable CDs and DVD-RAMs work.
Micron PCM uses a class of material alloy known as chalcogenides (“kal-koj--uh-nyde”). In basic terms, when electrical current and heat are applied, chalcogenides exhibit a reversible phase change phenomenon that can be used to store information through a resistance change in the material. More specifically, the heat causes the alloy to move from a highly unorganized, reflective, and resistive “amorphous” phase that does not store information well to a highly organized, low-resistance and reflectivity “polycrystalline” phase that does store information well. This results in reduced complexity and time-to-market for new designs as PCM breaks through architectural barriers and blends the best attributes of existing memory technologies into a single memory chip.
Advances in memory technology and pioneering work conducted by Micron has moved the technology to the forefront of the memory industry’s research and development activity. With exponential worldwide growth in electronic devices for the consumer, computer, and communication markets, PCM is a promising memory technology with the potential to enable new applications and memory architectures in a wide range of systems.