Unity reports progress on RRAM R&D trek

LONDON — Unity Semiconductor Corp., founded in 2002 by former Micron executive Darrell Rinerson, is a fabless chip company in the middle of a six-year R&D trek to bring a dense, fast nonvolatile memory to market. At the half-way point Rinerson is pleased with progress.

Rinerson’s chosen field of R&D is a two-terminal device known generically as Resistive RAM (RRAM), although as he pointed out in an interview with EE Times there are almost as many variants on the theme as there are companies pursuing this form of universal memory. And patent application record shows that RRAM has started to attract intense scrutiny from the likes of Samsung Electronics Co. Ltd., Sharp Corp. and Sony Corp. (see April 24 story).

“We [Unity] are at the stage of doing 64-kbit arrays. The next stage is multi-megabit arrays before coming out with gigabit commercial devices. We’re looking at two-and-a-half to three years before we launch a product. We’re being funded by patient, deep-pocketed investors,” said Rinerson

Rinerson, previously vice president of the Flash Memory Product Group at Micron Technology Inc. (Boise, Idaho) said Unity (Sunnyvale, Calif.) has received $16 million so far from August Capital, Morgenthaler and LightSpeed Ventures. The company has 26 employees.

RRAM development is generally based on using an electrical pulse to change the resistance across a thin film of either a simple binary metal oxide or a more complex perovskite oxide. It is the focus for several device manufacturers and some startups because it promises high density, low cost and low power consumption, along with non-volatility. It has the classic profile of the much-sought-after universal memory that could combine the speed of SRAM with density of DRAM and the nonvolatility of Flash memory. But there is risk because the basic physical mechanisms do not yet seem to be completely understood.

“It’s not surprising that it’s confusing because in some films multiple switching mechanisms are going on at the time,” said Rinerson. “But it is remarkably easy to get something to switch,” he added.

Rinerson said that the basic phenomenon had been known for thirty of forty years but because of gaps in the understanding it had been left unexploited. More recent work at the University of Houston and at IBM’s Zurich Labs in Switzerland had re-ignited interest among scientists and research engineers, said Rinerson.