IBM’S Millipede crawls toward replacement of flash memory

ZURICH, Switzerland — The latest announcement from IBM’s Zurich Research Laboratory about its progress in making memory devices based on MEMS and atomic force microscope tips hints at a leap in data storage capacity, but also highlights some of the hurdles still to be overcome.

The research team here, led by the distinguished duo of Peter Vettiger, an expert in microfabrication, and Gerd Binnig, who invented the scanning tunneling microscope, announced its first successful high-density storage operation using a single nanometer-sharp tip to bore 10 nanometer indentations or pits in a polymer surface (millionth of a millimeter) using heat to melt or displace the organic plastic.

The team working on what IBM calls the Millipede has an existing prototype with 1,024 legs, which it describes in detail along with the fabrication techniques in a paper published in the June 2002 inaugural issue of IEEE Transactions on Nanotechnology.

Click here to enlarge image

Now researchers want to multiply that capacity by four. Instead of just one atomic force microscope (ATM) probe, there will be more than 4,000 working in parallel on MEMS read/write heads operating over a patch that is 7 millimeters square, all controlled by a PC.

If it works, then the technology could be used in a memory chip with a data storage density of a trillion bits per square inch.

“Flash memory is not expected to surpass 1-2 gigabytes of capacity in the near term,” said the IBM researchers, “but Millipede technology could pack 10 to 15 gigabytes of data into the same tiny format, without requiring more power for device operation.”

The announcement marks a critical milestone on the way to IBM’s goal to replace flash memory devices. “Densities and yields of operation achieved with this first demo are very encouraging, although considerable improvements are possible in both areas,” write the researchers.

But there is a long list of steps to be undertaken before Millipede can be considered for commercial use, not least of which is an array four times the size of what exists today.

When the test of the next generation prototype takes place in spring 2003, the device won’t be writing at the terabit density proven this week. It will first see if it is possible build a 4,000-legged device. Vettiger said that at that stage he expects only small sections of the device to be tested at a time because of the limitations of the microcontroller circuits.

Market opportunity

Flash memory is used in digital cameras, MP3 devices, PDAs and mobile phones, a market where demand is strong. It is an elastic market too — as the price comes down the demand goes up,” said Steve Cullen, director and principal analyst at In-Stat/MDR, a high tech market research company.

The race is on to create a new generation of memory devices to improve upon and replace flash.

Carnegie Mellon University, Seagate Technology, Hewlett Packard and a number of startups, such as Nanochip Inc., are working on similar MEMS projects. Some researchers have the arrays move over a surface that contains bumps or dots representing data, while others are moving the surface below the atomic force probes, Cullen said.

The opportunity for Millipede and others like it would be to offer “significantly more storage capacity (than flash) at an acceptable cost or a major user benefit at virtually no price penalty”.

Can a device that contains nanosized legs by the thousands, nanotips and soft polymer surfaces be carried about in your backpack or pocket where MP3 players or mobile phones currently reside? It sounds too delicate.

Vettiger says it is not a problem. Very thin layers of silicon — 10 or 20 microns thick — are not as brittle as standard silicon chips. “At this thickness, you can bend it and twist it and it does not break.” He added that the bed that the polymer or plastic surface lies on is being equipped with a damping system made of a rubber material.

Currently, a team of eight people are working on the project, but there are others contributing from IBM Research labs elsewhere, such as the signal processing group next door in Zurich. Plus two groups in New York, one at the IBM T.J. Watson Research Center in Yorktown Heights, which is delivering servo/tracking and vibration know-how and the other at the IBM East Fishkill microelectronics center, which is developing the microcontrollers and new interconnection technologies for the chips.

Chemists at the IBM Almaden Research Center in California are tailoring existing polymers to the needs of this application. It’s a field of ongoing research. “Every polymer we’ve tested can handle the read/write indentations but they suffer mechanical wear. We are working on an enduring solution,” Vettiger said.

The involvement of so many research groups make clear the complexity of implementing a new category of memory device and IBM researchers are not hyping their progress. Only after a very long string of “ifs” and a pinch of good luck, does Vettiger offer his “best guess”.

“The earliest possible date for a commercial deployment in PDA, cameras and mobile phones of the memory cards based on Millipede technology is end of 2005,” he said.


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.