E-Paper Soon To Be in Living Color
Researchers have begun developing a color version of electronic paper, just when the monochromatic version has begun to take off in consumer applications. E Ink is one company working on the concept and has already released a demonstration version.
Color e-paper could be used in numerous applications, including e-books, e-magazines, tablet PCs, cellular telephones, and other handheld devices, said E Ink vice president of marketing Sriram K. Peruvemba.
It could also be used with billboards and other forms of outdoor advertising, said Lawrence Gasman, principal analyst at NanoMarkets, an industry-analysis firm.
E Ink has developed a demonstration version of color electronic paper, which could be used in applications like e-books, e-magazines, tablet PCs, cellular telephones, other handheld devices, and even outdoor advertising such as billboards. Several companies are working on color e-paper, just as consumers are beginning to adopt the monochromatic version. E-paper technology, around since the 1970s, uses millions of tiny microcapsules. In the monochromatic version, each microcapsule contains positively charged white and negatively charged black particles suspended in a clear fluid. If a negative electrical field is applied to a part of the area, the white particles move to the top of the display, and vice versa. This creates the desired display.
"There are many approaches to achieving color, and we plan to have a color product in 2010, so I don't want to divulge the approach we are taking at this point," said Peruvemba. "The demo units we have shown have a color filter on top of the display that is similar to what is used in LCD technology. Reflected light from monochrome e-paper passes through the color filter as needed to create the desired color images."
"And," he said, "we have an improved substrate material that increases contrast, renders the display brighter, and improves the speed of response [to commands to change images]." The company declined to identify the new material.
E Ink plans to use a flexible substrate for its thin-film-transistor displays so that users can roll them up and store or transport them more easily. Many companies—both manufacturers and device makers—have expressed interest in roll-up display applications. Currently, e-paper uses a film laminated on a glass substrate, which is rigid.
E Ink is not the only company working on color e-paper, said Gasman.
In general, he explained, the technology is not good enough for widespread commercial adoption yet. Moreover, he added, users have only now just begun adopting monochrome displays, although they eventually will demand color.
When products do come to market, he said, they will have to be high quality. The e-paper market cannot afford to bring a color product to market prematurely just to compete with other types of displays, he explained.
"It's going to be three or four years before you really see color good enough to go to market," he predicted.
News Briefs written by Linda Dailey Paulson, a freelance technology writer based in Ventura, California. Contact her at ldpaulson@yahoo.com.
IBM Develops a New Type of DNA Computing
Researchers are exploring new roles that DNA might play in computing. IBM scientists, along with California Institute of Technology senior research fellow Paul Rothemund, are assessing the feasibility of using the genetic material in self-assembly techniques for making semiconductors and other nanoscale devices.
The DNA would act as scaffolding for carefully arranging carbon nanotubes—strands of carbon atoms that conduct electricity—into arrays that could serve as chips for performing calculations or storing data. This differs from traditional DNA computing, in which molecules of the material use biological processes to perform computing-related tasks.
In the new approach, electron-beam lithography creates a pattern on the substrate with sites that bind DNA molecules in predetermined locations, said Greg Wallraff, an IBM Almaden Research Center research staff member.
The substrate is generally silicon coated with any of a number of films, said Jennifer Cha, IBM Almaden Research Center research staff member.
"The surface is covered by a water solution containing the DNA structures, which then self-assemble on the patterned features, one DNA molecule per site," Wallraff explained. The researchers then use the relatively large DNA molecules as templates for even smaller components such as carbon nanotubes or silicon nanowires," he said. The nanotubes or nanowires attach to the ends of the DNA strands in the pattern necessary to perform the desired tasks.
To be useful, self-assembling material must behave precisely and predictably, said Cha. Scientists are very familiar with DNA, which consists of specific chemical bases—such as guanine and cytosine—that behave dependably.
Because the technique works at nanoscale, IBM says, it could permit the manufacture of devices with circuit widths of one-eighth to one-tenth the 45 nanometers that traditional lithographic techniques can currently fabricate.
Smaller circuitry would let chip makers either pack more transistors onto processors, thereby making them more powerful, or make smaller chips with performance equal to today's larger versions.
The DNA-based process could occur in laboratories and thus could make chip manufacturing less expensive than current approaches, which require multibillion-dollar fabrication plants and expensive lithographic processing.
According to Cha, researchers are also working on ways to make the nanoscale wires that the new chips would need. In addition, she said, they want to improve yields and performance, as well as use DNA to separate and eliminate nanotubes that don't conduct electricity.
Moreover, scientists must improve the basic chip-making process to ensure that the nanotubes adhere to the DNA properly and in the correct orientation.
The technique may need 10 to 20 years of work before it can be used commercially, said Wallraff.
Researchers Use Software to Find Chip Flaws
Scientists have developed software that identifies problems in chips and recommends the best way to fix them. Developed at the University of Michigan, the software would help manufacturers cope with the growing number of bugs that take an increasing amount of time to fix as chips become more complex.
Processors now house a large number of transistors and perform a growing number of functions. This leads to more bugs and makes it difficult and time-consuming to identify them before vendors ship the chips, explained University of Michigan assistant professor Valeria Bertacco, who is working on the bug-finding software.
"As the complexity increases, so does the number of lines of code, which causes an explosion of bugs," said Gary Smith, chief analyst at the Gary Smith EDA consultancy.
This can delay the commercial release of chips and increase costs for manufacturers. Problems that occur after a processor ships can be particularly time-consuming and expensive to fix, Bertacco noted. Currently, though, fully debugging prototype chips can take up to a year, she said.
Because this would affect their budgets and market-related deadlines, chip makers usually fix only as many bugs as they can within a few months and then either ship their processors or cancel the project, according to Smith.
Manufacturers debug chips both before and after they are made, generally by designing and manufacturing special test boards and applying electric currents via the pins or internal nodes. Because of the analysis required to isolate and correct problems within the complex circuitry, the process relies on expensive logic analyzers to observe the internals.
The University of Michigan's FogClear software, run on an engineering workstation, could shorten the debugging process and reduce the number of prototypes and testing cycles vendors must conduct. It perhaps could even increase chips' reliability.
According to Bertacco, at any point in the chip-design flow, the software can examine two types of design-related errors: functional and electrical.
"A functional error is a bug in which the logic used to implement the design is incorrect in one or more circuit blocks," he explained. "An electrical error is one in which the circuit is functional but fails at the clock speed, voltage, and temperature intended for correct operation, typically because it doesn't finish evaluating within a clock cycle." Electrical errors tend to get worse as circuitry shrinks in size and performance demands grow.
The University of Michigan software uses mathematical techniques to examine the differences between the correct design and the actual circuit and find the precise location of bugs, said Bertacco. "Thus, the designer does not have to spend the effort required to wade through thousands of electrical inputs and millions of transistors to locate them."
"Our software can also narrow down the possible causes and develop fixes," he explained. "If we do not know that a chip has a bug, FogClear would not detect it by itself. However, when a bug has been observed, our software can locate and identify it."
The software conducts simulations of possible solutions to find the most cost-effective design variation that will fix the bugs, sometimes in ways that may be counterintuitive or not obvious to engineers doing the work in traditional ways.
In case studies, the researchers automatically repaired about 70 percent of major problems and reduced the debugging time from weeks to days, according to University of Michigan associate professor Igor Markov.
DOWNLOAD PDF 