Berthold Hahn | Meta: Which physical properties make microLEDs the superior choice for AI smart glasses over rival displays?
04:23 - 06:20
Other snippets from this talk
Summary of the clip:
Which physical properties make microLEDs the superior choice for AI smart glasses over rival displays?
The case starts with self-emission. For AI applications, there is normally not a lot of content: indicators and text coming back at the moment. A self-emissive display has a big advantage here. The material is inorganic, giving high stability, no burn-in risk, and lifetime degradation that is easy to model. Those traits let you build stable systems.
Speed is equally important. Nanosecond response is required because slow switching causes fatigue in daily-use glasses: motion blur follows the eye. Saccades create motion blur even when a viewer looks at a fixed image in the glasses, so displays must switch fast. Very high brightness supports readable text and indicators under those rapid eye movements. Fast, very high brightness displays are therefore needed here.
Brightness must also scale. The brightness is highly linear and very scalable. The system spans outdoors in the millinits range to indoors, where it must scale down several orders of magnitude. That requires very high stability of the light source and very good brightness scalability. Acceptance angle points to micron pitch; at 2.5-micron pitch, nominal emission is lower than a micron. Those tiny microns are why micro LED is judged best for these applications.
In this short video, you can learn:
* Self-emissive inorganic micro LED pixels offer high stability, no burn-in risk and easily modeled lifetime degradation.
* Nanosecond switching suppresses motion blur from saccades, even when the eye fixates on a static image in glasses.
* Highly linear, scalable brightness spans millinits outdoors to indoors across several orders of magnitude at micron-scale pitch.
📋 **Clip Abstract** The clip argues that micro LED suits AI smart glasses because it is self-emissive, inorganic and highly stable with no burn-in risk. It adds nanosecond response for motion blur and saccades, plus highly linear brightness scaling from millinits outdoors to indoors at 2.5-micron pitch.
About the speaker:
* Speaker: Berthold Hahn
* Company: Meta
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#SelfEmissiveMicroLED, #InorganicLifetimeStability, #NanosecondResponse, #MillinitBrightnessScaling, #ARDisplays, #SmartGlasses
This is a highlight of the presentation:
The Light Engine for the AI Glasses Era
More Highlights from the same talk.
02:18 - 03:46
Why must outdoor AR displays reach beyond 10 millinits in red, and why do phosphide LEDs work?
Why must outdoor AR displays reach beyond 10 millinits in red, and why do phosphide LEDs work?
Outdoor use is the hard case because applications must fight the sun's brightness. For text, the minimum contrast needed is about 1.6 to one. Indoors, the problem is manageable at about 200 lux. Outdoors, such as sports, getting information into a display runs down the battery very fast. That is why very high brightness and very good efficiencies are required to run these outdoor AR and VR displays.
With optical efficiencies of three to ten percent, you have to get into the millinits, especially for red, to have viable AR and VR outdoor displays. In current comparisons, micro LEDs are at about ten kilonits, while reading something can be about 200 kilonits. Quantum dot conversion is getting close to the millinit. Phosphide LEDs, however, definitely get beyond ten millinits. That puts them well into the range needed for high-efficiency glasses.
At the moment there is quite a gap between these technologies. There was always discussion about whether we have to go there, and whether phosphide is a possibility. What we see in the LEDs with phosphides is that it works. They definitely get beyond ten millinits, well into the range needed for high-efficiency glasses.
In this short video, you can learn:
* Outdoor text needs a minimum contrast of about 1.6 to one, while indoors about 200 lux is manageable.
* Optical efficiencies of three to ten percent require millinit-level output, especially red, for viable AR and VR outdoor displays.
* Quantum dot conversion is getting close to the millinit, but phosphide LEDs definitely get beyond ten millinits, showing the approach works.
📋 **Clip Abstract** Outdoor AR and VR displays must fight the sun, with text needing about 1.6 to one contrast and indoor operation manageable at about 200 lux. At three to ten percent optical efficiency, millinit red output is required; quantum dot conversion approaches it, but phosphide LEDs exceed ten millinits and work.
About the speaker:
* Speaker: Berthold Hahn
* Company: Meta
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#PhosphideLEDs, #QuantumDotConversion, #MillinitRed, #OutdoorARDisplay, #MicroLEDDisplays, #ARVRDisplays
13:55 - 15:14
How do LEDs compare to laser-based displays in coupling efficiency and waveguide integration?
How do LEDs compare to laser-based displays in coupling efficiency and waveguide integration?
At the moment, laser definitely has an advantage in coupling efficiency. That advantage also probably applies when comparing LEDs, where coupling remains the final frontier, especially as lower pixel sizes drive more complex designs. From physics, the laser has much better directionality. So for reaching toward 100 million pieces, the laser's directional output is the fundamental benefit, while LEDs must solve how to get there.
Laser systems, especially when working with waveguides, are a challenge. Laser has better coupling efficiency, but waveguide integration is difficult. If you can solve this one and get the cost down accordingly, there could be a path forward. Laser systems with waveguides remain a challenge, and solving that with cost reduction is the condition for a path forward.
At the moment, at least for the next years, LEDs will dominate this. The later lasers come, the more difficult it will be to get into this market. Laser has a physics advantage in directionality and coupling efficiency, but LEDs have the near-term position. The ongoing discussion—LEDs versus lasers—turns on waveguide integration and cost, not only coupling efficiency.
In this short video, you can learn:
* Laser-based displays currently have a clear advantage in coupling efficiency, and physics gives lasers much better directionality.
* Laser systems working with waveguides are a challenge; solving that and lowering cost could create a path forward.
* For at least the next years LEDs will dominate, and the later lasers arrive, the harder it will be to enter the market.
📋 **Clip Abstract** The clip compares laser-based displays with LEDs on coupling efficiency and waveguide integration. It concludes that lasers have better directionality and current coupling efficiency, but waveguide challenges and cost must be solved, while LEDs are expected to dominate for at least the next years.
About the speaker:
* Speaker: Berthold Hahn
* Company: Meta
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#LaserBasedDisplays, #CouplingEfficiency, #WaveguideIntegration, #LEDDominance, #AROptics, #DisplayLightEngines
07:19 - 09:46
How do phosphide, quantum dot conversion, and native InGaN red microLEDs compare in brightness, lifetime, and wavelength shift?
How do phosphide, quantum dot conversion, and native InGaN red microLEDs compare in brightness, lifetime, and wavelength shift?
For red microLEDs, phosphides are mature and scalable. The material system reaches 10 millinits, with pitches lower than two microns and a basis at sub-one micron. Homogeneity of arrays is solved, yields are very good, and even high homogeneity is achievable. Results now enter the millinit range. This scalable red approach enables wide field of view doors, outdoor displays, and is moving into product delivery.
Quantum dot conversion reaches 500K to one millinit with a usable lifetime; that lifetime is the main limitation. Pushing very high brightness creates a trade-off in lifetime. This path also opens indoor AR and some indicator glasses with more field of view, where it is probably bright enough. It is not a one-fits-all solution; the best value proposition must be chosen per use case.
Native red InGaN remains promising. It allows monolithic integration, as colleagues from Google are doing, but top brightness is still limited. Even with very high efficiencies, peak luminance at long wavelength and the considerable wavelength shift of nitrides are a massive killer for outdoor high-brightness applications. The way to see it: it is not one-fits-all; fit the best value proposition for each one.
In this short video, you can learn:
* Phosphide red microLEDs are mature, scalable below one micron, with solved array homogeneity and very good yields at 10 millinits.
* Quantum dot conversion delivers 500K to one millinit with usable lifetime, but brightness increases trade against lifetime.
* Native red InGaN enables monolithic integration but suffers limited top brightness and considerable wavelength shift at long wavelengths, harming outdoor high-brightness use.
📋 **Clip Abstract** Red microLEDs split into phosphides, quantum dot conversion and native red InGaN, each with distinct brightness, lifetime and wavelength-shift limits. Phosphides offer mature submicron scalability and outdoor capability, quantum dots trade brightness for usable lifetime and open indoor AR, while native InGaN promises monolithic integration but faces top-brightness and wavelength-shift barriers.
About the speaker:
* Speaker: Berthold Hahn
* Company: Meta
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#PhosphideMicroLEDs, #QuantumDotConversion, #NativeInGaNRed, #WavelengthShift, #MicroLEDDisplays, #ARDisplays