Power/Performance Bits: March 31

Reusable gallium arsenide wafers; to tackle integer overflow, software debug adopts hardware techniques; micro-magnets are batteries for bacteria.


Reusable gallium arsenide wafers

A manufacturing process developed by Stanford researchers could dramatically reduce the cost of gallium arsenide electronics, potentially opening up new applications for the material.

In the search for silicon’s replacement, gallium arsenide (GaAs) has much to offer on performance. It’s faster than silicon, less noise, and features a wide direct band gap—which makes it particularly appealing for solar. “Solar cells that use gallium arsenide hold the record when it comes to the efficiency at which they convert sunlight into electricity,” said Bruce Clemens, Stanford professor of materials science and engineering.

On price point, however, GaAs doesn’t fare quite so well: It can cost about $5,000 to make a wafer of gallium arsenide 8 inches in diameter, versus $5 for a silicon wafer, according to Aneesh Nainani, who teaches semiconductor manufacturing at Stanford.

To lessen this thousand-to-one cost differential, the team set out to make gallium arsenide wafers reusable.

A brief overview of the Stanford process (Source: Clemens Research Group)

The process, documented in the journal MRS Communications, relies on a disposable indium gallium arsenide nitride layer grown between the substrate and the gallium arsenide film. After a standard gas deposition process, the researchers adhered a flexible polymer substrate to the GaAs film before vaporizing the sacrificial layer with an IR laser.

After a thorough cleaning, the gallium arsenide wafer was ready to make the next batch of circuits.

Nainani estimates that this reuse could create gallium arsenide devices that would be 50 to 100 times more expensive than silicon circuits – still a big difference, but much less than the thousand-fold increase that exists today.

For integer overflow, software debug adopts hardware techniques

Integer overflows are one of the most common bugs in computer programs — not only causing programs to crash but, even worse, potentially offering points of attack for malicious hackers.

Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory will present a new algorithm for identifying integer-overflow bugs at this month’s ACM International Conference on Architectural Support for Programming Languages and Operating Systems. The researchers tested the algorithm on five common open-source programs, in which previous analyses had found three bugs. The new algorithm found the known bugs—and 11 new ones.

Their system, dubbed DIODE (for Directed Integer Overflow Detection), begins by feeding the program a single sample input. As that input is processed, however, the system records each of the operations performed on it by adding new terms to what’s known as a symbolic expression.

When the program reaches a point at which an integer is involved in a potentially dangerous operation, like a memory allocation, DIODE records the current state of the symbolic expression. The initial test input won’t trigger an overflow, but DIODE can analyze the symbolic expression to calculate an input that will.

“DIODE is based on symbolic execution, a state-of-the-art technique that provides the ability to automatically explore and analyze paths through a program by modeling these paths as mathematical formulas,” said Cristian Cadar, senior lecturer in computing at Imperial College London.

This is an example of techniques that have been used in the hardware world for a long time now being deployed in the software world. Interest in this type of deployment is one of the reasons Synopsys made their recent acquisition of Coverity and why other EDA vendors are looking at ways technologies developed for EDA can be utilized in other industries.

Micro-magnets are batteries for bacteria

Humans aren’t the only ones to use batteries—it seems microorganisms take advantage of rechargeable energy too. Researchers from the University of Tübingen, the University of Manchester, and the Pacific Northwest National Laboratory discovered that certain bacteria can use tiny magnets in order to share electrons, storing and using it just as we use a rechargeable battery.

The experiment focused on two types of bacteria that metabolize iron to generate energy. The team incubated the microbes with magnetite, a magnetic nanoparticle found in many soils and sediments, then controlled the amount of light the cultures were exposed to. In conditions replicating day, iron-oxidizing bacteria removed electrons from the magnetite, discharging it. During night-time conditions, the iron-reducing bacteria took over and were able to dump electrons back onto the magnetite and recharge it for the following cycle.

Bottle used in the experiment, with pink color due to the Fe(II)-oxidizer bacteria and magnetite clinging to the side of the bottle near a magnet. (Source: James Byrne/Tübingen)

Bottle used in the experiment, with pink color due to the Fe(II)-oxidizer bacteria and magnetite clinging to the side of the bottle near a magnet. (Source: James Byrne/Tübingen)

This oxidation/reduction mechanism continued through several day/night cycles, showing the battery was used repeatedly. While this work has been on iron-metabolizing bacteria, the team thinks the potential for magnetite to act as a battery could extend to bacteria which don’t normally require iron to grow.

“This may have some interesting geochemical applications. There has been considerable recent work on using magnetite to clean up toxic metals,” said Andreas Kappler, professor at Tübingen and secretary of the European Association of Geochemistry. “For example, magnetite can reduce the toxic form of chromium, chromium VI, to the less toxic chromium (III), which can then be incorporated into a magnetite crystal. The fact that this magnetite may then be exposed to these reducing bacteria could potentially enhance its remediation capacity. But we are still at an early stage of understanding the bioengineering implications of this discovery.”

Leave a Reply

(Note: This name will be displayed publicly)