Intel Claims Phase-Change Memory Breakthrough

Intel and Numonyx has said research engineers have found a way to stack up 64MB single-layer PCM (phase-change memory) arrays within a single die into what the companies describe as a true “cross-point array” containing both memory and a switch for interconnection purposes.

The result is a vertically integrated memory cell called a PCMS (phase-change memory and switch). A PCMS consists of one PCM element layered with an OTS (Ovonic Threshold Switch).

This promises substantial gains in scalability, performance and power savings that are projected to be magnitudes better than what is possible in current solid-state and spinning disk storage, said Al Fazio, Intel fellow and director of memory technology development.

A PRAM (phase-change memory) chip is nonvolatile memory that is works equally well for executing code and storing large amounts of data, giving it capabilities of both flash memory and DRAM (dynamic RAM). This means PCM can execute code with performance, store larger amounts of memory and sustain millions of read/write cycles.

Stacked up, these chips have the potential for read/write speed and storage capability far beyond anything yet seen in electronics.

The key descriptor here, however, is the word “potential.” Fazio and Numonyx Senior Technology Fellow Greg Atwood explained that such a multilayer PCM processor has not actually been tested: Only the single-layer 64MB chip has been tested, and the stackability attributes have been identified and designed.

“Our testing has involved single-layer 64MB PCM array on top of a CMOS. We’ve tested the array for reliability and distributional properties,” Fazio told eWEEK during a conference call. “The multiple-layer version—whether it’s two or four or multiple layers stacked up on top of each other—is yet to be done. But this is the first and most significant milestone, which is the basic stacking in a large, multimegabit array.”

A whitepaper with details on the multilayer PCM breakthrough—and more information about the performance of a stacked PCM array—will be made available in December at the International Electron Devices Meeting in Baltimore, Fazio said.

The breakthrough, Fazio said, was producing the building blocks and demonstrating the PCM chip at a single layer.

“The first layer is the hardest layer,” Atwood told eWEEK.

Overall, the results of more than six years of PCM research are extremely promising, Atwood said.

“The results show the potential for higher density, scalable arrays and NAND-like usage models for PCM products in the future,” he said. “This is important as traditional flash memory technologies face certain physical limits and reliability issues, yet demand for memory continues to rise in everything from mobile phones to data centres.”

PCM, which stands to replace DRAM and other components in digital devices during the next five to 10 years, offers a great deal of promise for the entire IT industry. It is blazingly fast, reliable and versatile. However, it is expensive—about 10 times the cost of DRAM—and in the current economic climate it has been a tough sell as a replacement technology.

The pricing, however, is expected to come down over time as fabricating processes are improved.

Intel debuted its first PCM chips at its developers’ conference in San Francisco in September 2006. The wafer shown to eWEEK that day represented Intel and STMicroelectronics’ first grasp of the new type of nonvolatile memory chip. Intel later started up a new division and merged it with STM to create Numonyx.

A great deal of development has been completed at Numonyx since then; Numonyx CTO Ed Doller told eWEEK on 11 Aug. that adoption by mainstream IT companies has been slow but that it would take use by only a couple of big names—Apple and Microsoft would be two of them—for PCM technology to take off into the market stratosphere.

“PCM is on the verge, and we think it’s inevitable that it will replace a lot of what is in devices today,” Doller said. “But to start, it would take a company that is good at producing both hardware and software to make best use of it. It’s bound to happen.”

Doller said he thinks PCM will need three to five more years to attain widespread adoption.

PCM chips use the same material, chalcogenide, that is used inside to store data on rewritable optical disks. But instead of using a laser to change the properties of the material and thus create the zeros and ones that make up data, the chips use electricity that flows through a resistor. The resistor heats up and does the job of the laser, changing the materials’ properties to represent a zero or a one.

The phase-change process has been used mostly in rewritable DVDs and CDs, but bringing it to the enterprise level for data center use has been a difficult process.

The effort is the “culmination of [work by] some of the smartest materials guys on the planet,” Doller said. “Over the years, this has an opportunity to be a very large memory technology.”

Most industry people and analysts insist that PCM has the potential to replace both NAND—flash memory designed primarily for data storage—and NOR flash memory, designed for executing code, with one type of chip, streamlining manufacturing processes.