Skip to Content

Low-Power Transistors May Boost Wearable Computer Battery Life

Small, portable devices could get new stamina thanks to a transistor design that can cut a computer chip’s power consumption in half.
September 10, 2013

A new way of designing chips could solve one of the biggest problems facing wearable computers such as Google Glass and the Samsung smart watch—their batteries generally have to be recharged every day.

Efficiency inside: This Fujitsu chip for processing images in digital cameras is the first in production based on a new transistor design that offers significant power savings.

The novel design comes from SuVolta, which has been working since 2006 to dramatically improve the energy efficiency of transistors, the fundamental component of computer chips. The company has received $62 million in venture funding. At at the industry conference Hot Chips in California last month, SuVolta showed results from an experiment in which its technology was used to make a version of an existing chip. SuVolta’s version consumed half the power of the original while running at the same speed. It could operate 35 percent faster than the conventional chip if consuming the same power.

SuVolta’s technology isn’t suited to making very high-performance processors, but the company’s chief technology officer, Scott Thompson, says it is a good match for the battery-powered computers most popular today. “Computing is no longer about making $200 CPUs,” Thompson says, referring to the processors at the heart of traditional PCs. “It’s about $100 cell phones and wearable devices.”

The chip that got an upgrade from SuVolta was the Cortex M0, designed by ARM Holdings, whose designs underpin the processors used in most mobile devices, including those of Apple and Samsung (see “Moore’s Law Is Becoming Irrelevant”). The M0 is a simple chip used by ARM as a benchmark to test the potential for more powerful designs, says SuVolta’s senior director of digital design, David Kidd, who led the M0 experiment.

SuVolta says it is working with chip maker United Microelectronics, based in Taiwan, on a manufacturing process that could be used for mobile chips. And last week, Fujitsu announced that it has begun mass producing an image-processing chip destined for digital cameras using SuVolta’s technology. Thompson says his company has four other undisclosed partnerships with chip manufacturers.

SuVolta’s technology saves power by reducing the variability in the performance of different transistors on the same chip. The variations stem from minor fluctuations in the material a chip is made from and cause some transistors to require higher voltages than others to operate properly. That wastes power because a chip’s overall voltage must be set at a level high enough for all transistors to work, causing many to get more than they need.

SuVolta’s design reduces the variation between transistors by modifying the “channel” of a transistor, which either allows or blocks the flow of current as a transistor switches on and off. By adding trace amounts of compounds that tweak silicon’s electrical properties, SuVolta splits a transistor’s channel into three layers. The electrical interplay of those three layers makes transistors largely resistant to the causes of variation in the voltages they need to work.

SuVolta’s new transistor design can be made by existing manufacturing lines without major changes. That’s in contrast with a more complex, 3-D transistor design called the FinFET, pioneered by Intel and also being adopted by its competitors (see “3-D Transistors”). Chip fabrication plants require major refits to handle that technology.

However, SuVolta’s technology will not increase performance as much as FinFET transistors do. Mike Demler, an analyst with the Linley Group, a semiconductor research firm, believes that will limit the inroads that SuVolta’s technology will make in mobile devices. “The transistor will work, but you will always get a slower circuit compared to FinFETs,” he says.

SuVolta’s Thompson counters that there still aren’t any cell phones on the market with FinFET-based chips and that it will be a long time before they appear in other, fast-growing areas of mobile and compact computing. “No one’s talking about using FinFETs in the smart watch devices recently announced by Samsung and Qualcomm, or for the Internet of Things,” he says.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.