Calit2 : NVM Workshop

NVM Workshop

More than 50 academic and industry researchers attended the 2001 Workshop on Non-Volatile Memory, sponsored by Calit². A blue-chip roster of speakers from leading companies in the information storage industry delivered presentations on September 21 and 22 at the main auditorium of the Center for Magnetic Recording Research at UCSD's Jacobs School of Engineering. (For a detailed rundown of the topics discussed, see the Summary of Proceedings.)

The audience included 11 faculty members who participate in the Materials & Devices group within Calit²-seven from UCSD, four from UC Irvine. "The institute recognizes the importance of putting academics and business researchers together," said Frances Hellman, organizer of the workshop and a professor of physics at UCSD. "It quickly became clear that there are a number of areas where our research can complement what industry is doing, especially on technologies that won't be commercial for a few more years." (Click here for Q&A with Hellman.)

In welcoming workshop participants, Calit² director Larry Smarr said that the institute's buildings at UC Irvine and UC San Diego are being designed to enhance research in key areas for materials and devices. "Opto-electronics, quantum computing and communications, non-volatile memories, and materials research are all core competencies for Calit²," said Smarr. "Non-volatile memory has a major role to play in a future where billions of wireless sensors and other devices will be gathering, storing, transmitting and even processing vast amounts of information."

UCSD physics professor Ivan Schuller updated attendees on basic research issues related to the future of magnetics, urging scientists to "follow your nose! Don't let applications or immediate business dictate where you look for the answers." Schuller also advised researchers to "push limits"-to smaller sizes, more complex materials, using state of the art technologies, and exploring unusual geometries. Looking out ten years, the leader of Calit²'s materials and devices group at UCSD (his counterpart at UC Irvine is professor G.P. Li) said the technology frontier will involve self assembly, nanomagnetism, and "spintronics" (spin electronics). Added Schuller: "If you look 20 years into the future, you're talking about quantum magnetics, working at the molecular level-and quantum computing."

After Schuller's presentation, a senior industry executive expanded on "The Future of Magnetic Recording." "Magnetic recording will not be dislodged in the foreseeable future," said Dieter Weller, director of media research at Pittsburgh-based Seagate Technology. "It is a very low-cost method of storing vast amounts of data with almost no competition. Magnetic recording is hard to compete with from a capacity perspective as well as from a technology maturity perspective. At the moment, recording densities of about 100 Gigabits per square inch have been demonstrated. We think we can push the technology up to perhaps 150-200 Gigabits per square inch using conventional longitudinal [technology], then we may move to perpendicular recording pushing toward 500 Gigabits per square inch or higher. At terabit per square inch areal densities a bit cell is only 25 nano-meters on the side. We are eventually approaching the point where we have single-, as opposed to multiple-particle-per-bit recording. At Seagate we are very interested in novel media approaches. One example is self-assembly of magnetic nano-particles with length scales in the 3-10 nm range. Writing and reading such small magnetic islands, however, will be extremely challenging. We need a lot of innovation. Eventually, some kind of probing technology will be required: it could be based on spintronics effects, tunneling mechanisms, magnetic force detection or perhaps advanced high sensitivity giant-magneto-resistance sensors."

Several groups have proposed MEMS based probe storage devices involving an x-y media stage and an array of individually actuated probe tips. Such devices could provide faster access to data than conventional rotating magnetic storage, yet provide large capacities at relatively low cost. They would be intermediate in the storage hierarchy, bridging the space between the fast but expensive DRAM and the massive and low cost magnetic storage. The media could be non-magnetic, e.g. ferroelectric or polymer materials. "At Seagate Research, we are interested in all these areas," added Weller. "Our philosophy is to place 'some' effort inside our own research laboratory, where researchers work on various approaches, materials and phenomena and stay in touch with what's happening on the outside. The further out we go, the more we have to rely on partners and academia." Seagate already sponsors fellowships and Ph.D. theses at various institutions all over the country, but is not a partner in Calit². "Not yet," said Weller. "But I'm sure things could be arranged."

Also making presentations at the September 21 session were two industry researchers who focused on MRAM development: NVE's Jim Daughton, on "MRAM: Status and Opportunities"; and Motorola's Brad Engel, on "Prototype 256-kbit MRAM." The workshop concluded September 22 with presentations on "Magnetic Tunneling Junctions and Magnetic Memory" (by Stuart Parkin of IBM Almaden); Joachim Ahner (Seagate Technology) on "Novel Imaging Technology and Potential Multiple Probe Storage Devices"; and "Ferro-electric Memories: Science, Technology and Beyond," by Ramamoorthy Ramesh (University of Maryland).

After the workshop, organizer Frances Hellman was upbeat about the prospects of future cooperation between UC faculty, students and industrial partners. "We will now work on a research agenda to meet some of the challenges outlined by industry researchers," said Hellman. "We will also explore with them the possibility of internships for students, and we may also invite key scientists back to the campus to participate in Calit²'s ongoing series of short courses."

Newsroom > Web Articles

NVM Workshop