Skip to Content

New Memory Chips Store Data Not with Electricity, but with Light

Researchers demonstrate new photonic memory that keeps working even when the power is off.
September 29, 2015

Computing companies searching for more speed have started to use light to transport data inside computers instead of electricity. Now researchers have unveiled a promising scheme for using light to store information on a chip as well—even when the power is off.

In this electron microscope image, a tiny patch of a phase-change material called GST (yellow) sits on a waveguide, serving as a memory cell that can be written and read by sending light pulses through the guide.

Using light instead of electricity to move information between a computer’s memory and its processor could lead to much faster and more energy-efficient computers (see “Intel’s Laser Chips Could Make Data Centers Run Better”). But right now it is necessary to convert the optical signals to electrical ones and store the data electronically, which is relatively slow compared with the speed of today’s processors. The new “all-photonic” memory, which takes advantage of the same materials used in rewritable CDs and DVDs, is a step toward systems that achieve more efficient data transfer and storage, according to the technology’s inventors.

Photonic memory has been demonstrated on a chip before, but it was short-lived and it required a constant supply of light to work. This is the first “on-chip” optical memory that is nonvolatile, meaning that it does not require a constant supply of energy, and thus can provide long-term storage the way a hard drive can. The basis of the technology is a so-called phase-change material. Light pulses can be used to switch the material between two distinct states—one in which the atoms are ordered, or crystalline, and one in which they are disordered, or amorphous. The researchers exploited this phenomenon to write and read information.

One particular attribute of this material makes it especially useful for memory storage. The researchers showed they could use light to put the material into mixed states—say, 10 percent crystalline and 90 percent amorphous, or 20 percent crystalline and 80 percent amorphous, and so on. Having more than just two states available for memory storage “means you can cram a lot more information” into the same space, says Harish Bhaskaran, a professor of materials science and a nanoengineering expert at the University of Oxford, in the U.K. Bhaskaran, together with Wolfram Pernice, of the University of Münster, in Germany, led the research. 

In the near term, a memory technology like this could be used to augment the performance of data centers and thus expand the kinds of applications possible thanks to cloud computing. Several major computing companies are developing systems for moving light around a chip using waveguides, or from one chip to another using optical cables like those common in the telecommunications industry. Bhaskaran and his colleagues say the new memory scheme is compatible with conventional optical fibers, as well as waveguides.

The technology is still far from commercialization. The researchers only demonstrated the ability to read and write several bits. More research and development will be needed to understand how exactly it can or should be applied.

One avenue Bhaskaran and Pernice plan to explore is the design of unconventional computer architectures, perhaps including ones meant to emulate the way brains processes information, that might overcome fundamental speed and efficiency limits of traditional electronic computers (see “Thinking in Silicon”). Bhaskaran says the same technique they used to exploit multiple states of the phase-change material for memory storage can be used to perform basic arithmetic operations, like counting. “If you can do sequential counting, then you can do computation.”

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.