The February 2023 issue of IEEE Spectrum is here!

Close bar

Spring-loaded Pixels to Drive Holographic Displays

Semiconductor research firm Imec hopes to create holographic video displays using MEMS

3 min read
Spring-loaded Pixels to Drive Holographic Displays

Earlier this month, I sat through a 3-D showing of Martin Scorsese’s new movie, Hugo. I was surprised to find myself distracted by discomfort. Every twenty minutes or so I had to close my eyes to give them a rest. By the end of the film, I had a big headache. I’m not the only one who’s run into this problem. Stereoscopic projection, the sort that’s standard in cinemas, is hard on the brain. The main culprit, according to researchers and cinematographers, is that the illusion creates conflict between two key processes in vision. While our eyes focus on the plane of the screen, they must orient themselves to converge between the screen and the viewer.

Holographic video, which is capable of creating virtual images that are truly three-dimensional, has long been a tantalizing alternative to today’s 3-D display technology. But it's been moving forward in fits and starts. Every once in a while, a new development in holographic display will make the headlines. But the real quarry truly immersive video  is still a far-off dream. After years of work, the best holographic video displays are still quite bulky, slow, and blurry.

Take the news in January out of MIT’s Object-Based Media Group, which rigged a Kinect camera to capture three-dimensional data at a record-setting rate of 15 frames per second. That’s fast for holographic video capture, about half of what’s needed to create the perception of continuous motion on a television set. But the resulting holographic images, which were made by sending light through tunable crystals, didn’t really complete the loop. I have tried, but I must admit I can’t really see Princess Leia at all in the team’s recent demo.

A group at the European research firm Imec suspects semiconductors could come to the rescue. Researchers there have hatched a scheme to construct holographic displays by shining laser light on microelectromechanical systems (MEMS) platforms capable of moving up and down like small, reflective pistons. Each pixel, which would have a spring-like mechanism attached to it, could be pulled down or released by applying or removing a voltage.

Making moving holograms

When I visited Imec headquarters in Leuven, Belgium in October, the firm was showing off the very first stage of this technology, a MEMS-less chip containing a fixed image of Imec’s new research building. The chips were made by growing a layer of silicon oxide onto silicon wafer and then selectively etching away square patches of the silicon oxide to create a pre-programmed, checkerboard-like pattern in which etched-away pixels were about 150 nanometers or so lower than their neighbors. The entire chip was then topped with a reflective coating of aluminum.

When laser light shines on the chip, it bounces off of the boundary between adjacent pixels at an angle. Across the chip, this diffracted light interferes constructively and destructively to create a 3-D picture.

The team hopes to make these holographic pictures move by swapping fixed structures for moving pixels – small mirrored platforms that can move up and down many times a second to create a moving projection.

Doing this is no small feat, says Richard Stahl, senior researcher of Imec’s research group on vision systems. To get the diffraction needed to make a holographic image, each MEMS structure must be smaller than the wavelength of light that shines on the chip, which in this case means each structure must be about 0.5 by 0.5 micron-wide, less than a tenth the size of the MEMS structures the firm currently makes. Each pixel must also be able to move up or down by a hundred or so nanometers without colliding with its neighbors.

By building the structures out of a mix of silicon and germanium, which the company has already used to make tiltable MEMS mirrors, Imec expects to be able to integrate the data-handling logic needed to control the structures directly beneath the pixels themselves. That should make it relatively easy to make fast displays, Stahl says. The team hopes to construct the first, proof-of-concept moving structures beginning around the middle of 2012, with the hope of graduating on to constructing entire chips tiled with the MEMS structures.

Even if these sorts of chips can be made, the cost of replacing a theater projection system with a screen full of MEMS will likely keep this technology small for quite some time. But when I visited Imec, my guide, program manager Francesco Pessolano, told me we could be about 5 years away from letter-size displays containing about 400 billion moving pixels. I conducted a short (2-D) video interview with Pessolano when I visited Leuven. Here he is describing the technology:

Note: I visited Imec headquarters during the company’s annual forum for technology reporters. My travel expenses were partly subsidized by the company.

The Conversation (0)

Two Startups Are Bringing Fiber to the Processor

Avicena’s blue microLEDs are the dark horse in a race with Ayar Labs’ laser-based system

5 min read
Diffuse blue light shines from a patterned surface through a ring. A blue cable leads away from it.

Avicena’s microLED chiplets could one day link all the CPUs in a computer cluster together.

Avicena

If a CPU in Seoul sends a byte of data to a processor in Prague, the information covers most of the distance as light, zipping along with no resistance. But put both those processors on the same motherboard, and they’ll need to communicate over energy-sapping copper, which slow the communication speeds possible within computers. Two Silicon Valley startups, Avicena and Ayar Labs, are doing something about that longstanding limit. If they succeed in their attempts to finally bring optical fiber all the way to the processor, it might not just accelerate computing—it might also remake it.

Both companies are developing fiber-connected chiplets, small chips meant to share a high-bandwidth connection with CPUs and other data-hungry silicon in a shared package. They are each ramping up production in 2023, though it may be a couple of years before we see a computer on the market with either product.

Keep Reading ↓Show less