Click below to download the white paper

In Part 1 of this series, we compared the processes, performance and features of glass waveguides fabricated using Myrias Optics’  Nano Imprint Lithography (NIL) approach using all-inorganic ink vs. an all-etched fabrication approach. Part 1 showed that an all-inorganic NIL waveguide can match the performance of an etched solution for high refractive index glass.

In Part 2, released here, we show the results of a rigorous cost model analysis of the two approaches to achieve 1 million yielded waveguides. The bottom line – there is a significant cost advantage with the Myrias Optics’ NIL approach compared to the etch approach. At a $350 substrate cost for example, 3-etch waveguides are 3.4 times more expensive to process. 

Why is the etch process so much more expensive? While costs for masters/photomasks, substrates, ink/resist, facilities and labor vary in a range of dollars or tens of dollars, the high capital expense of the semiconductor etch equipment is the main driver. The depreciation expense for the 3-etch process for example, is 21 times higher than the Myrias Optics’ depreciation expense – a significant contributor to final waveguide cost.

Part 1 noted that the direct-etch approach can, in principle, scale to meet 100-million-unit annual targets, but that doing so requires proportionally larger capital equipment (CapEx) commitments — the model’s findings here make that concrete. Because the etch process cost structure is dominated by depreciation rather than substrate or labor, its cost per waveguide does not improve meaningfully with scale in the way a substrate, or labor, dominated cost structure might; it simply requires a proportionally larger multiple of an already tool-heavy production line.

Myrias Optics’ Type 2 route has a substrate-driven cost structure that behaves differently at scale: substrate cost per wafer is largely scale-invariant, so the primary lever for further cost reduction at high volume is substrate sourcing and yield improvement rather than incremental tool purchases.

Meta has put forth an industry goal to make 100 million waveguides that will allow a monocular display-based AR optional addition to ordinary eyewear at a $100 retail price point. This $100 add-on must include manufacturing margins for a display, optics, electronics and a waveguide at a minimum, suggesting the waveguide price should cost less than $35 (with profit margin) at high volume.