Erhan Ercan | Morphotonics: If 100 million AR glasses ship, how many 300mm wafer fabs must run just for waveguides at 90% yield?
02:38 - 03:35
Other snippets from this talk
Summary of the clip:
If 100 million AR glasses ship, how many 300mm wafer fabs must run just for waveguides at 90% yield?
Cost is key β it is what makes the big difference at the end of the day. Take a slide from the referenced presentation: assume 24 waveguides on a 300-millimeter wafer. Run that at 90% yields and you get 22 waveguides from a single 12-inch wafer. That is the arithmetic the whole scaling argument rests on.
Scale it to 100 million glasses and the number stops being academic: that volume equals the capacity of 60 SMCs. The presenter is explicit that he has not seen whether 24 waveguides per 300-millimeter wafer is real β the figure is offered as a hypothetical. Either way, the demand case implies dozens of fabs' worth of output.
What has worked before is larger glass. That is what took flat panel displays from so expensive to far more affordable at 65 inch and beyond. As glass size expands, the industry is now at gen 10, gen 10 and a half, and that larger substrate really helped the scaling route the display industry has taken. The question is whether waveguide manufacturing can borrow that playbook.
In this short video, you can learn:
* At 90% yield, 24 waveguides per 300-millimeter wafer delivers 22 waveguides per 12-inch wafer.
* 100 million AR glasses would consume the capacity of 60 SMCs on those assumptions.
* Flat panel displays scaled by moving to larger glass, now at gen 10 and gen 10 and a half.
π **Clip Abstract** Cost is framed as the decisive factor in AR waveguide manufacturing. At 24 waveguides per 300-millimeter wafer and 90% yield, 100 million glasses implies the capacity of 60 SMCs, while the display industry's move to larger glass β now gen 10 and gen 10.5 β stands as the scaling precedent.
About the speaker:
* Speaker: Erhan Ercan
* Company: Morphotonics
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#300mmWafer, #ARWaveguideYield, #Gen10Glass, #12InchWafer, #AROptics, #ARDisplays
This is a highlight of the presentation:
Multi-Wafer Nanoimprint Lithography: A Panel-Based Approach to Scalable AR Waveguide Manufacturing
More Highlights from the same talk.
05:22 - 06:39
What feature range and stamp reuse have roll-to-plate nanoimprint demonstrated in real production?
What feature range and stamp reuse have roll-to-plate nanoimprint demonstrated in real production?
Roll-to-plate is the patterning route discussed. It scales from gen three and gen three-and-a-half up to gen five size imprints, which reaches 1.1 by 1.3 meters. That panel-scale format is the core manufacturing proposition: larger imprinted areas without moving to roll-to-roll or injection molding. The demonstrated capability is not limited to one structure class; it covers a broad feature range and has been shown in production hardware, not only laboratory coupons.
Replicated structures range from 50 nanometers up to 500 microns in size, and the process has gone even below that. Track pitch variation has reached the picometer range, which is not something people usually expect from nanoimprint. This combination matters where nanoscale features and large panel areas must coexist. The claim is based on actual replication results, not a single hero demo.
In production, a booth monitor is a Samsung display with a 2D/3D switchable function. Its lenticular lens was made on these tools. The key manufacturing proof is stamp lifetime: over 1,000 times of working stamp reuse has been shown. That reuse level is central to panel-based nanoimprint economics, because it links mastered nanoscale features to repeated large-area replication without replacing the stamp after every imprint.
In this short video, you can learn:
* Roll-to-plate imprints scale from gen three and three-and-a-half up to gen five, reaching 1.1 by 1.3 meters.
* Replication covers 50 nanometers to 500 microns and has demonstrated picometer-range track pitch variation.
* A production 2D/3D switchable Samsung monitor uses a lenticular lens made on these tools, with over 1,000 working stamp reuses.
π **Clip Abstract** Roll-to-plate nanoimprint scales panel imprints from gen three and three-and-a-half up to gen five, at 1.1 by 1.3 meters. It replicates features from 50 nanometers to 500 microns and below, shows picometer-range track pitch variation, and has achieved over 1,000 working stamp reuses on a production lenticular lens for a 2D/3D switchable Samsung monitor.
About the speaker:
* Speaker: Erhan Ercan
* Company: Morphotonics
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#RollToPlateNanoimprint, #WorkingStampReuse, #PicometerTrackPitch, #SwitchableLenticularLens, #NanoimprintLithography, #PanelScalePatterning
13:08 - 14:15
When will high-index glass arrive in 300mm-plus panels, and what polishing challenges remain?
When will high-index glass arrive in 300mm-plus panels, and what polishing challenges remain?
About four years ago, working with SCHOTT, the first-ever 300 millimeter squared, 1.9 index glass was demonstrated and shown at SPIE. These glasses are not made in round format; they are cored. The challenge becomes how to polish them to achieve the right TDE warp and bow. That creates a chicken-and-egg situation for adoption.
Last week, SCHOTT announced 310 by 310, 2.0 index with lightweight. Corning, Hoya and others have made glasses in other formats as well. The industry is pushing that way, but the question remains: do you need it? Right now, a lot of production is happening only at eight-inch. It is not even at 12-inch. The view is that this is the path forward.
The substrate question includes HRI and silicon carbide at larger packages, essentially four-inch. The answer: larger glass formats are arriving, but production maturity lags. Announced 300 millimeter squared and 310 by 310 panels do not mean volume at 12-inch. Polishing to achieve TDE warp and bow remains the practical hurdle. The industry must decide whether demand justifies the larger panel path.
In this short video, you can learn:
* A 300 millimeter squared, 1.9 index glass was demonstrated with SCHOTT about four years ago and shown at SPIE.
* These high-index glasses are cored rather than made in round format, so polishing to the right TDE warp and bow is the challenge.
* SCHOTT recently announced 310 by 310, 2.0 index lightweight glass, yet a lot of production is still only at eight-inch, not 12-inch.
π **Clip Abstract** Larger high-index glass formats are being announced, including 300 millimeter squared 1.9 index and 310 by 310 2.0 index lightweight. The remaining barriers are polishing to achieve TDE warp and bow, plus the chicken-and-egg question of whether production demand justifies moving beyond eight-inch.
About the speaker:
* Speaker: Erhan Ercan
* Company: Morphotonics
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#HighIndexGlass, #SiliconCarbideSubstrates, #TDEWarpBow, #310x310Glass, #ARWaveguides, #DisplaySubstrates
07:31 - 09:06
How does a multi-wafer carrier turn four 300mm wafers into one panel to hit 60 UPH?
How does a multi-wafer carrier turn four 300mm wafers into one panel to hit 60 UPH?
Cypress is a multi-wafer carrier tool, so you are not bound to one wafer at a time. It processes four 12-inch, four 300 millimetre squared, or even four 310 by 310 squared substrates, or nine 8-inch substrates simultaneously. The system sees the whole thing as one big panel. There is no high-index glass at that size; if there were, life would have been much easier.
The tool is modular. Substrates go through a primer for spray coating, then a bake station for pre- and post-bake, then an inkjet station where resin is dispensed, and finally into the nanoimprint photography module where the actual final imprinting happens. That sequence is the system, built now for an alpha customer rather than as a paper exercise.
Throughput is the point: 15 carriers per hour, which equates to about 60 UPH for 12-inch and about 135 for 8-inch. Because the system treats multiple wafers as one big panel, the multi-wafer carrier is what turns four 300 millimetre wafers into a single panel run. Cypress will be available by the end of this year.
In this short video, you can learn:
* Cypress is a multi-wafer carrier tool that processes four 12-inch, four 300 mm squared, four 310 by 310 squared or nine 8-inch substrates at the same time, treating them as one big panel.
* The modular system runs a primer for spray coating, a pre- and post-bake station, an inkjet station that dispenses resin, then the nanoimprint photography module for final imprinting.
* The tool has a capacity of 15 carriers per hour, equating to about 60 UPH for 12-inch and about 135 for 8-inch substrates.
π **Clip Abstract** Multi-wafer nanoimprint lithography uses a carrier that treats several substrates as one panel, so the tool is not bound to processing one wafer at a time. The modular Cypress line moves substrates through primer spray coating, pre- and post-bake, inkjet resin dispensing and nanoimprint, reaching 15 carriers per hour or roughly 60 UPH for 12-inch and 135 for 8-inch.
About the speaker:
* Speaker: Erhan Ercan
* Company: Morphotonics
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#MultiWaferCarrier, #NanoimprintModule, #InkjetResinDispensing, #CarriersPerHour, #ARWaveguides, #SemiconductorManufacturing