Co-Packaged Optics (CPO) is usually described as a distance problem. Move the optical engine closer to the switch ASIC or XPU. Shorten the electrical path. Cut SerDes power. Raise bandwidth density.
That direction is real. Optical functions are gradually moving closer to compute. But shorter distance does not eliminate physics. A physical bottleneck rarely disappears; it simply moves to another place in the system.
As electrical reach shrinks, more responsibility shifts into the package: substrate geometry, EIC/PIC integration, optical alignment, thermal and mechanical stability, manufacturing, test, and reliability. That raises a larger question: what physical platform can let an increasingly diverse set of electronic and photonic functions coexist around very large compute devices?
An increasingly interesting answer may be glass.
The opportunity is not merely replacing organic substrates with glass, or replacing silicon interposers with glass. Glass may eventually become a scalable heterogeneous integration platform around the compute chip for both electronic and photonic functions. That possibility deserves more attention.