A TDK research team has developed a photonic integrated circuit for laser-scanning displays in smart glasses, demonstrating a 10GHz modulation bandwidth and full-color output. SPIE, the international society for optics and photonics, described the work on October 19, 2026, which was published in the journal Advanced Photonics Nexus. Unlike the conventional approach of rapidly varying the current driving the laser, the technology controls the laser light after it has been emitted, using a thin-film lithium niobate chip. The aim is to achieve both higher resolution and lower power consumption in compact displays. However, the performance of this optical circuit should be kept distinct from the display performance of a finished pair of smart glasses.
Separating laser emission from brightness control
A laser-scanning display shows images by moving laser light at high speed and drawing pixels in sequence. The system combines red, green, and blue laser light into a single beam and adjusts the brightness of each color to match the scanning position. To display high-resolution images, it is not enough to move the mirror quickly; the laser intensity must also be switched rapidly in step with that movement.
Conventional systems adjust brightness through "direct modulation," varying the current supplied to each color's laser diode. According to SPIE's explanation, however, direct modulation has a more limited modulation bandwidth than methods that use an external optical modulator. In addition, a laser needs current above a certain threshold to oscillate, so controlling it directly through current has posed challenges for both speed and power consumption.
The photonic integrated circuit developed by TDK's Yasuhiro Takagi and colleagues handles this brightness control in a device separate from the laser source. The material is thin-film lithium niobate, whose optical properties change when voltage is applied. The circuit incorporates three Mach-Zehnder interferometers, one each for red, green, and blue. These split light into two paths and then recombine it, using the resulting interference to adjust light intensity, which allows the brightness of each color to be controlled at high speed by voltage. Finally, a waveguide-based combiner merges the three colors into a single beam.
With this configuration, the roles of generating light and varying its intensity to match the image are separated, so the laser itself no longer needs to be driven at high speed. The team controlled each color independently with voltage and successfully produced full-color output. They also combined the circuit with a MEMS scanning mirror, which moves a tiny mirror, showing its potential for laser-scanning displays.
10GHz bandwidth and a reported power reduction of more than 75%
The 10GHz figure confirmed in this research refers to the frequency bandwidth of the optical modulator, which changes light intensity at high speed. The research team says this is sufficient bandwidth to support high-resolution video at 4K and above. That does not mean, however, that 4K video has been successfully shown on finished smart glasses. The results SPIE describes concern measurement of the modulation bandwidth, generation of full-color output, and combination with a MEMS scanning mirror.
To raise the resolution of a laser-scanning display, the scanning position and the brightness of each color must be precisely synchronized as the number of drawn pixels grows. Being able to switch light intensity quickly removes one factor that hinders higher resolution, but actual display performance also depends on the scanning mirror's speed, the optical design, and other elements. The fact that this circuit has the bandwidth needed for 4K or higher is therefore a separate stage from achieving equivalent resolution in a display device that incorporates it.
The research also reported a notable result for power consumption. According to SPIE, the team reported that overall system power consumption could be cut by more than 75% compared with conventional methods. Because brightness is controlled by an external optical modulator, the laser's drive current no longer has to change rapidly, which improves power efficiency. The researchers also say that neutral-density filters for dimming the light become unnecessary, reducing the number of optical components.
That said, "overall system" should not be read as the entire smart glasses product. SPIE's explanation does not disclose the absolute power consumption before and after the reduction, nor details of the comparison conditions such as display brightness. It does not mean that the power consumption of the whole wearable device, including communications and processing circuitry for image handling, fell by more than 75%, and this figure alone does not indicate how much battery life would improve.
The findings are reported in the paper "Thin-film lithium niobate photonic integrated circuit for laser beam scanning imaging" (DOI: 10.1117/1.APN.5.6.066007), published in Advanced Photonics Nexus, Volume 5, Issue 6.
Development continuing from the 2024 video-display demonstration
TDK has been developing technology that uses voltage to control the brightness of visible-light lasers for some time. In the technology TDK announced on October 9, 2024, the company said it could be more than 10 times faster than conventional visible-light laser control and could support video resolutions of 4K and above. It also demonstrated video display with a thin-film lithium niobate device combined with QD Laser's direct retinal drawing technology.
Whereas the 2024 announcement described low power consumption as an expected benefit, SPIE's explanation of the new work gives a specific reduction figure of more than 75%. It cannot be confirmed, however, whether the prototype used in 2024 and the optical circuit evaluated in this research are the same sample. The 75% figure should therefore not be taken as an improvement over the 2024 prototype; it should be treated as the reduction this research reported in comparison with conventional methods.
On manufacturing, TDK explained as early as 2024 that it formed the lithium niobate thin film by sputtering. Sputtering is a deposition technique that lays material down in a thin layer on a substrate, and the company positions it as a manufacturing method suited to mass production. This suggests TDK is working on fabrication methods with future volume production in mind, not only on high-speed optical modulation. Suitability for mass production, however, is not the same as being able to begin mass production.
TDK is also developing other optical technologies for smart glasses. In an official announcement on October 2, 2026, the company said it had successfully demonstrated a retinal-drawing display using a "meta-optics mirror" that is 150nm thick and has 80% visible-light transmittance. The ultra-thin mirror can be built into ordinary eyeglass lenses and aims to give a more natural appearance than methods using conventional curved mirrors. Under a business cooperation agreement with QD Laser, TDK is developing a next-generation optical engine that combines a full-color laser module, light control devices, and the meta-optics mirror.
The meta-optics mirror demonstration announced on October 2 is, for now, monochrome, and TDK aims to support full color and begin mass production within several years. It is a different technology from the photonic integrated circuit for which full-color output was confirmed in this research, and a finished product combining the two has not yet demonstrated full-color video display.
As future challenges, SPIE cites reducing propagation loss within the waveguides and combining loss when merging red, green, and blue light into one beam. If light lost inside the optical circuit can be suppressed, the efficiency of the whole optical system could improve further, potentially contributing to lower power consumption. The work improves the performance of an important component of smart glasses optics, but commercialization will require evaluation of the light source, modulator, scanning mirror, and projection optics together. The key question going forward is how far power consumption and optical loss can be reduced while maintaining practical display brightness and resolution.